Forum adverts like this one are shown to any user who is not logged in. Join us by filling out a tiny 3 field form and you will get your own, free, dakka user account which gives a good range of benefits to you:
No adverts like this in the forums anymore.
Times and dates in your local timezone.
Full tracking of what you have read so you can skip to your first unread post, easily see what has changed since you last logged in, and easily see what is new at a glance.
Email notifications for threads you want to watch closely.
Being a part of the oldest wargaming community on the net.
If you are already a member then feel free to login now.
Marmatag wrote: Screening, Holding objectives, psychic denial, these are all more valuable than people give them credit for. And I don't agree in regards to the minimum guard. Look at the top 5 BAO.
Like i said you will see every once in a while someone will bring slightly more guard but that doesnt take away from the actual stats of the tournament. Lists that brought majority guard had a lower win percentage and points per round then list that brought minimum batteries at BAO. This was then shown to be effective once again at nova where guard was not the majority of the tp lists
LOL you are talking about people bringing GAKING BRIGADES man. This isn't "MIN GAURD" Min gaurd is a battalion with no mortars.
Minimum Guard is the minimum required in a list to achieve what you need - if that is 5CP (+ regen), then a minimum battalion is what you need. If you need 12CP (+ regeneration), then a minimum Brigade is what you need to achieve minimum Guard.
Whether the number of armies fielding either as a soup ingredient at the next big event changes is, currently, anyone's guess...
Gamgee on Tau Players wrote:we all kill cats and sell our own families to the devil and eat live puppies.
Kanluwen wrote: This is, emphatically, why I will continue suggesting nuking Guard and starting over again. It's a legacy army that needs to be rebooted with a new focal point.
Confirmation of why no-one should listen to Kanluwen when it comes to the IG - he doesn't want the IG, he want's Kan's New Model Army...
tneva82 wrote: You aren't even trying ty pretend for honest arqument. Open bad faith trolling.
- No reason to keep this here, unless people want to use it for something...
Asmodios wrote: A. You would need to make the strategems of equal power for both army A and B for this to become valid.
Yes you would and should! The idea that some armies have better stratagems by design rather than accident has absolutely no evidence. Sure, some armies have better stratagems, but some armies have better units too. It is not because GW intentionally tried to make it so, they just failed at balancing.
Also not sure how you can not agree that soup is inherently more powerful....... look at the number of soup armies in the top of tournaments and the number of mono guard...... there's a reason why all the top players for all the top factions are bringing soup and not mono guard lists. Take one look at nova and explain to me how soup isn't the issue 10/10 of the top lists are soup
"Yes, there is some extreme soup builds that are better than most (or even all) monobuilds, but I'd argue that most soup combinations are actually weaker than mono guard. If you punish soup in general while leaving guard as it is, then it just widens that gap. "
What part of this you didn't understand? Tournaments are representative of the absolutely cutting edge builds, and sure those seem to be some specific soup builds (mainly due poor balancing between codices.) It doesn't change the fact that excluding those couple of anomalies, mono guard is stronger than most soup builds. Do you think Black Templars + Sisters is stronger than mono-Guard? Ad Mech + Inquisition stronger than mono-Guard? No, they aren't. Sure Salamanders + Guard is stronger than pure Salamanders, but that's because Guard is disproportionately strong, and again pure Guard list would still be stronger.
Xenomancers wrote: If that is true (i'd like to see the source) but what does that mean if the winning undefeated list are sporting IG brigades as their primary and often include more IG than the minimum? It clearly can't mean that putting more IG in your list is bad for your win rate.
First time using chess clocks at a gt/major on day 2 and despite leaning towards them not being the definitive answer before the event, I’ll admit I’m now firmly a supporter. Even in my last game when we had an issue with our clock working it was easy to call the judge over for a fix and it was flawless from there. Biggest benefit to them was that it keeps both players much more involved in the game imo. Doing so immediately gets the game moving and keeps it on track for a natural conclusion. None of my games were called to time. Round one: natural conclusion turn 4 – win
Round two: natural conclusion turn 3 – loss
Round three: natural conclusion turn 5 – win
Round four: natural conclusion turn 6 – win
Round five: natural conclusion turn 5 – loss
Round six: natural conclusion turn 6 – win
3 hour rounds were perfect. Holy crap does that extra few minutes help complete everything. In the past I’ve almost always felt rushed to complete my game, log the results and then run off to the next game with a stress filled running late feeling. Not the case with these round times. Adding 45 total minutes to the day was completely worth it imo.
So, we’ve got in the top 10:
Death Guard x 1
Imperial Knights x 2
Astra Militarum x 3
Custodes x 1
Thousand Sons x 1
Eldar/Ynnari x 2
Pretty interesting top 10 and a nice spread. It gets even more diverse when we look at the top 20:
Death Guard x 1
Imperial Knights x 3
Astra Militarum x 3
Custodes x 1
Thousand Sons x 2
Eldar/Ynnari x 3
Dark Angels x 1
T’au x 2
Drukhari x 1
Cult Mechanicus x 1
Tyranids x 2
Some great data to look over from the event in this sheet, here.
Fun facts:
Most common Primary Factions:
Imperial Knights: 14
T’au: 13
Astra Militarum: 11
Asuryani: 11
Tyranids: 10
Most common Secondary Detachments:
Astra militarum: 40
T’au: 20
Asuryani: 19
Drukhari: 15
Tyranids: 12
Most points earned per round:
Adeptus Sororitas: 27.34
Dark Angels: 26.50
Imperial Knights: 24.12
Renegade Knights: 24
Ynnari: 23.85
Highest win percentage:
Renegade Knights: 75%
Dark Angels: 75:
Genestealer Cults: 58.33%
Drukhari: 55.32%
Imperial Knights: 55.28%
T’au: 55.07%
found it for you. Taking IG as your primary detachment meant on average you didnt not make the top for Win % or most points per round
Automatically Appended Next Post:
Marmatag wrote: The problem is you can't look at the field and draw a conclusion in regards to balance. And, i've seen that data, it did not account for "percentage used," it just had if a faction was used, at all. You are flatly misrepresenting this data. I would suggest you provide what you have.
You have to filter out the players who lose all, or most of their games. There's a reason people look at the lists that finish in the top 20 or so. Because those lists performed well. I don't care if B. Billy Bumshoes brought a "fluffy" list and got stomped out. That doesn't mean anything in regards to game balance. And the bottom of any tournament has a ton of total crap lists, especially if it's a big event.
You are actively arguing that we should induce survivorship bias into our statistical analysis
Asmodios wrote: A. You would need to make the strategems of equal power for both army A and B for this to become valid.
Yes you would and should! The idea that some armies have better stratagems by design rather than accident has absolutely no evidence. Sure, some armies have better stratagems, but some armies have better units too. It is not because GW intentionally tried to make it so, they just failed at balancing.
Also not sure how you can not agree that soup is inherently more powerful....... look at the number of soup armies in the top of tournaments and the number of mono guard...... there's a reason why all the top players for all the top factions are bringing soup and not mono guard lists. Take one look at nova and explain to me how soup isn't the issue 10/10 of the top lists are soup
"Yes, there is some extreme soup builds that are better than most (or even all) monobuilds, but I'd argue that most soup combinations are actually weaker than mono guard. If you punish soup in general while leaving guard as it is, then it just widens that gap. "
What part of this you didn't understand? Tournaments are representative of the absolutely cutting edge builds, and sure those seem to be some specific soup builds (mainly due poor balancing between codices.) It doesn't change the fact that excluding those couple of anomalies, mono guard is stronger than most soup builds. Do you think Black Templars + Sisters is stronger than mono-Guard? Ad Mech + Inquisition stronger than mono-Guard? No, they aren't. Sure Salamanders + Guard is stronger than pure Salamanders, but that's because Guard is disproportionately strong, and again pure Guard list would still be stronger.
What you are arguing is that mono guard is stronger then soup because of soup builds that we dont see and have no data on. Using that logic i could say that mono guard is awful because everyone we aren't looking at is spamming chimeras and sentinals. We have to work within the data data provided or we are simple just stating opinion with nothing to back up our claims
This message was edited 3 times. Last update was at 2018/10/04 16:56:30
According the previous list i posted, the maximum number of CPs at 2000 points with 3 detachments and no regeneration for an IG list that wants to bring a Castellan, is 20 CP.
Do you think that these CPs would be enough to fuel both the castellan and the IG? Do you think it could be competitive?
Honestly i don't see it winning anything serious. Folds to Aeldari and to anything that can shutdown the Castellan. You are basically banking on the firepower of the Castellan, which even with the Raven buff is still quite randomic.
Brigade, Battallion, Super-Heavy Det with 2 Armigers and the Castellan = 23 cp. I won’t even bother debunking the rest of the nonsense in this thread when you can’t even get the most rudimentary facts straight.
Good riddance!
If that is your attitude then you don't have anything useful to add to the discussion.
Marmatag wrote: Screening, Holding objectives, psychic denial, these are all more valuable than people give them credit for. And I don't agree in regards to the minimum guard. Look at the top 5 BAO.
Like i said you will see every once in a while someone will bring slightly more guard but that doesnt take away from the actual stats of the tournament. Lists that brought majority guard had a lower win percentage and points per round then list that brought minimum batteries at BAO. This was then shown to be effective once again at nova where guard was not the majority of the tp lists
LOL you are talking about people bringing GAKING BRIGADES man. This isn't "MIN GAURD" Min gaurd is a battalion with no mortars.
Minimum Guard is the minimum required in a list to achieve what you need - if that is 5CP (+ regen), then a minimum battalion is what you need. If you need 12CP (+ regeneration), then a minimum Brigade is what you need to achieve minimum Guard.
Whether the number of armies fielding either as a soup ingredient at the next big event changes is, currently, anyone's guess...
Hummm - so don't we just call this min maxing? Which is what you do at tournaments.
I'd argue that A brigade of anything is not "minimum" Nor did anyone take a min brigade. They could have taken even cheaper ones. They took the most effective brigade they could and were willing to put extra points into it.
The way I see it? I need lots of CP - so min guard is not an option - I have to go for LOTS of guard to do this. AND IT WORKS.
This message was edited 1 time. Last update was at 2018/10/04 18:57:53
If we fail to anticipate the unforeseen or expect the unexpected in a universe of infinite possibilities, we may find ourselves at the mercy of anyone or anything that cannot be programmed, categorized or easily referenced.
- Fox Mulder
Xenomancers wrote: If that is true (i'd like to see the source) but what does that mean if the winning undefeated list are sporting IG brigades as their primary and often include more IG than the minimum? It clearly can't mean that putting more IG in your list is bad for your win rate.
First time using chess clocks at a gt/major on day 2 and despite leaning towards them not being the definitive answer before the event, I’ll admit I’m now firmly a supporter. Even in my last game when we had an issue with our clock working it was easy to call the judge over for a fix and it was flawless from there. Biggest benefit to them was that it keeps both players much more involved in the game imo. Doing so immediately gets the game moving and keeps it on track for a natural conclusion. None of my games were called to time. Round one: natural conclusion turn 4 – win
Round two: natural conclusion turn 3 – loss
Round three: natural conclusion turn 5 – win
Round four: natural conclusion turn 6 – win
Round five: natural conclusion turn 5 – loss
Round six: natural conclusion turn 6 – win
3 hour rounds were perfect. Holy crap does that extra few minutes help complete everything. In the past I’ve almost always felt rushed to complete my game, log the results and then run off to the next game with a stress filled running late feeling. Not the case with these round times. Adding 45 total minutes to the day was completely worth it imo.
So, we’ve got in the top 10:
Death Guard x 1
Imperial Knights x 2
Astra Militarum x 3
Custodes x 1
Thousand Sons x 1
Eldar/Ynnari x 2
Pretty interesting top 10 and a nice spread. It gets even more diverse when we look at the top 20:
Death Guard x 1
Imperial Knights x 3
Astra Militarum x 3
Custodes x 1
Thousand Sons x 2
Eldar/Ynnari x 3
Dark Angels x 1
T’au x 2
Drukhari x 1
Cult Mechanicus x 1
Tyranids x 2
Some great data to look over from the event in this sheet, here.
Fun facts:
Most common Primary Factions:
Imperial Knights: 14
T’au: 13
Astra Militarum: 11
Asuryani: 11
Tyranids: 10
Most common Secondary Detachments:
Astra militarum: 40
T’au: 20
Asuryani: 19
Drukhari: 15
Tyranids: 12
Most points earned per round:
Adeptus Sororitas: 27.34
Dark Angels: 26.50
Imperial Knights: 24.12
Renegade Knights: 24
Ynnari: 23.85
Highest win percentage:
Renegade Knights: 75%
Dark Angels: 75:
Genestealer Cults: 58.33%
Drukhari: 55.32%
Imperial Knights: 55.28%
T’au: 55.07%
found it for you. Taking IG as your primary detachment meant on average you didnt not make the top for Win % or most points per round
Automatically Appended Next Post:
Marmatag wrote: The problem is you can't look at the field and draw a conclusion in regards to balance. And, i've seen that data, it did not account for "percentage used," it just had if a faction was used, at all. You are flatly misrepresenting this data. I would suggest you provide what you have.
You have to filter out the players who lose all, or most of their games. There's a reason people look at the lists that finish in the top 20 or so. Because those lists performed well. I don't care if B. Billy Bumshoes brought a "fluffy" list and got stomped out. That doesn't mean anything in regards to game balance. And the bottom of any tournament has a ton of total crap lists, especially if it's a big event.
You are actively arguing that we should induce survivorship bias into our statistical analysis
Asmodios wrote: A. You would need to make the strategems of equal power for both army A and B for this to become valid.
Yes you would and should! The idea that some armies have better stratagems by design rather than accident has absolutely no evidence. Sure, some armies have better stratagems, but some armies have better units too. It is not because GW intentionally tried to make it so, they just failed at balancing.
Also not sure how you can not agree that soup is inherently more powerful....... look at the number of soup armies in the top of tournaments and the number of mono guard...... there's a reason why all the top players for all the top factions are bringing soup and not mono guard lists. Take one look at nova and explain to me how soup isn't the issue 10/10 of the top lists are soup
"Yes, there is some extreme soup builds that are better than most (or even all) monobuilds, but I'd argue that most soup combinations are actually weaker than mono guard. If you punish soup in general while leaving guard as it is, then it just widens that gap. "
What part of this you didn't understand? Tournaments are representative of the absolutely cutting edge builds, and sure those seem to be some specific soup builds (mainly due poor balancing between codices.) It doesn't change the fact that excluding those couple of anomalies, mono guard is stronger than most soup builds. Do you think Black Templars + Sisters is stronger than mono-Guard? Ad Mech + Inquisition stronger than mono-Guard? No, they aren't. Sure Salamanders + Guard is stronger than pure Salamanders, but that's because Guard is disproportionately strong, and again pure Guard list would still be stronger.
What you are arguing is that mono guard is stronger then soup because of soup builds that we dont see and have no data on. Using that logic i could say that mono guard is awful because everyone we aren't looking at is spamming chimeras and sentinals. We have to work within the data data provided or we are simple just stating opinion with nothing to back up our claims
Nothing interesting about that top 10. It's exactly expected. 3 AM/ 2 Knights. Every single list including knights and AM some with BA - some with Custodes. A few Ynnari. A few Choas soupers. This is exactly expected. Also - top 10 has 3 AM primary...yet - you are trying to make a case that it hurts your win % to do this? Are you arguing in good faith here seriously?
This message was edited 1 time. Last update was at 2018/10/04 18:56:14
If we fail to anticipate the unforeseen or expect the unexpected in a universe of infinite possibilities, we may find ourselves at the mercy of anyone or anything that cannot be programmed, categorized or easily referenced.
- Fox Mulder
Automatically Appended Next Post: Still no non-knifeear xenos in the top 10. That didn't change.
No breakdown on the soup involved in the post above. I'm betting they're mostly not pure, but is there a breakdown of total faction presence in the top 10 (not just 'primary')?
This message was edited 1 time. Last update was at 2018/10/04 18:57:15
Xenomancers wrote: If that is true (i'd like to see the source) but what does that mean if the winning undefeated list are sporting IG brigades as their primary and often include more IG than the minimum? It clearly can't mean that putting more IG in your list is bad for your win rate.
First time using chess clocks at a gt/major on day 2 and despite leaning towards them not being the definitive answer before the event, I’ll admit I’m now firmly a supporter. Even in my last game when we had an issue with our clock working it was easy to call the judge over for a fix and it was flawless from there. Biggest benefit to them was that it keeps both players much more involved in the game imo. Doing so immediately gets the game moving and keeps it on track for a natural conclusion. None of my games were called to time. Round one: natural conclusion turn 4 – win
Round two: natural conclusion turn 3 – loss
Round three: natural conclusion turn 5 – win
Round four: natural conclusion turn 6 – win
Round five: natural conclusion turn 5 – loss
Round six: natural conclusion turn 6 – win
3 hour rounds were perfect. Holy crap does that extra few minutes help complete everything. In the past I’ve almost always felt rushed to complete my game, log the results and then run off to the next game with a stress filled running late feeling. Not the case with these round times. Adding 45 total minutes to the day was completely worth it imo.
So, we’ve got in the top 10:
Death Guard x 1
Imperial Knights x 2
Astra Militarum x 3
Custodes x 1
Thousand Sons x 1
Eldar/Ynnari x 2
Pretty interesting top 10 and a nice spread. It gets even more diverse when we look at the top 20:
Death Guard x 1
Imperial Knights x 3
Astra Militarum x 3
Custodes x 1
Thousand Sons x 2
Eldar/Ynnari x 3
Dark Angels x 1
T’au x 2
Drukhari x 1
Cult Mechanicus x 1
Tyranids x 2
Some great data to look over from the event in this sheet, here.
Fun facts:
Most common Primary Factions:
Imperial Knights: 14
T’au: 13
Astra Militarum: 11
Asuryani: 11
Tyranids: 10
Most common Secondary Detachments:
Astra militarum: 40
T’au: 20
Asuryani: 19
Drukhari: 15
Tyranids: 12
Most points earned per round:
Adeptus Sororitas: 27.34
Dark Angels: 26.50
Imperial Knights: 24.12
Renegade Knights: 24
Ynnari: 23.85
Highest win percentage:
Renegade Knights: 75%
Dark Angels: 75:
Genestealer Cults: 58.33%
Drukhari: 55.32%
Imperial Knights: 55.28%
T’au: 55.07%
found it for you. Taking IG as your primary detachment meant on average you didnt not make the top for Win % or most points per round
Automatically Appended Next Post:
Marmatag wrote: The problem is you can't look at the field and draw a conclusion in regards to balance. And, i've seen that data, it did not account for "percentage used," it just had if a faction was used, at all. You are flatly misrepresenting this data. I would suggest you provide what you have.
You have to filter out the players who lose all, or most of their games. There's a reason people look at the lists that finish in the top 20 or so. Because those lists performed well. I don't care if B. Billy Bumshoes brought a "fluffy" list and got stomped out. That doesn't mean anything in regards to game balance. And the bottom of any tournament has a ton of total crap lists, especially if it's a big event.
You are actively arguing that we should induce survivorship bias into our statistical analysis
Asmodios wrote: A. You would need to make the strategems of equal power for both army A and B for this to become valid.
Yes you would and should! The idea that some armies have better stratagems by design rather than accident has absolutely no evidence. Sure, some armies have better stratagems, but some armies have better units too. It is not because GW intentionally tried to make it so, they just failed at balancing.
Also not sure how you can not agree that soup is inherently more powerful....... look at the number of soup armies in the top of tournaments and the number of mono guard...... there's a reason why all the top players for all the top factions are bringing soup and not mono guard lists. Take one look at nova and explain to me how soup isn't the issue 10/10 of the top lists are soup
"Yes, there is some extreme soup builds that are better than most (or even all) monobuilds, but I'd argue that most soup combinations are actually weaker than mono guard. If you punish soup in general while leaving guard as it is, then it just widens that gap. "
What part of this you didn't understand? Tournaments are representative of the absolutely cutting edge builds, and sure those seem to be some specific soup builds (mainly due poor balancing between codices.) It doesn't change the fact that excluding those couple of anomalies, mono guard is stronger than most soup builds. Do you think Black Templars + Sisters is stronger than mono-Guard? Ad Mech + Inquisition stronger than mono-Guard? No, they aren't. Sure Salamanders + Guard is stronger than pure Salamanders, but that's because Guard is disproportionately strong, and again pure Guard list would still be stronger.
What you are arguing is that mono guard is stronger then soup because of soup builds that we dont see and have no data on. Using that logic i could say that mono guard is awful because everyone we aren't looking at is spamming chimeras and sentinals. We have to work within the data data provided or we are simple just stating opinion with nothing to back up our claims
Nothing interesting about that top 10. It's exactly expected. 3 AM/ 2 Knights. Every single list including knights and AM some with BA - some with Custodes. A few Ynnari. A few Choas soupers. This is exactly expected. Also - top 10 has 3 AM primary...yet - you are trying to make a case that it hurts your win % to do this? Are you arguing in good faith here seriously?
Notice that primary guard did not make the top spots for win percentage or points earned per round. That primary guard preformed bellow mono armies like Tau. primary guard is stronger then mono guard and still finished bellow must take mono armies. This kind of evidence supports my previous claims that wile mono guard is good it is actually not supperior to armies like tau that strangely dont have 3 threads a day about how they need a nerf.
Marmatag wrote: The problem is you can't look at the field and draw a conclusion in regards to balance. And, i've seen that data, it did not account for "percentage used," it just had if a faction was used, at all. You are flatly misrepresenting this data. I would suggest you provide what you have.
You have to filter out the players who lose all, or most of their games. There's a reason people look at the lists that finish in the top 20 or so. Because those lists performed well. I don't care if B. Billy Bumshoes brought a "fluffy" list and got stomped out. That doesn't mean anything in regards to game balance. And the bottom of any tournament has a ton of total crap lists, especially if it's a big event.
You are actively arguing that we should induce survivorship bias into our statistical analysis
You have absolutely no idea what you're talking about. Unless you're prepared to argue that the quality of lists and players is the same regardless of win/loss, you have no point whatsoever. My friend recently went 0-3 with Guard + Knights at an RTT. He forgot his models had invulnerable saves. As you can imagine, he's super casual and just enjoys the game. Should he count for the purposes of determining game balance?
You should actually go to a tournament. The bigger the tournament, the more casual players are there. They get weeded out early. The honest truth is if you are a competitive player and you consider yourself skilled, you should be 3-0 (at least 2-1) going into day 2 in any big event. It's unlikely you'll draw a nightmare matchup day 1 with a top 10 player.
This message was edited 2 times. Last update was at 2018/10/04 19:26:35
Galas wrote: I remember when Marmatag was a nooby, all shiney and full of joy. How playing the unbalanced mess of Warhammer40k in a ultra-competitive meta has changed you
Bharring wrote: He'll actually *change his mind* in the presence of sufficient/sufficiently defended information. Heretic.
Xenomancers wrote: If that is true (i'd like to see the source) but what does that mean if the winning undefeated list are sporting IG brigades as their primary and often include more IG than the minimum? It clearly can't mean that putting more IG in your list is bad for your win rate.
First time using chess clocks at a gt/major on day 2 and despite leaning towards them not being the definitive answer before the event, I’ll admit I’m now firmly a supporter. Even in my last game when we had an issue with our clock working it was easy to call the judge over for a fix and it was flawless from there. Biggest benefit to them was that it keeps both players much more involved in the game imo. Doing so immediately gets the game moving and keeps it on track for a natural conclusion. None of my games were called to time. Round one: natural conclusion turn 4 – win
Round two: natural conclusion turn 3 – loss
Round three: natural conclusion turn 5 – win
Round four: natural conclusion turn 6 – win
Round five: natural conclusion turn 5 – loss
Round six: natural conclusion turn 6 – win
3 hour rounds were perfect. Holy crap does that extra few minutes help complete everything. In the past I’ve almost always felt rushed to complete my game, log the results and then run off to the next game with a stress filled running late feeling. Not the case with these round times. Adding 45 total minutes to the day was completely worth it imo.
So, we’ve got in the top 10:
Death Guard x 1
Imperial Knights x 2
Astra Militarum x 3
Custodes x 1
Thousand Sons x 1
Eldar/Ynnari x 2
Pretty interesting top 10 and a nice spread. It gets even more diverse when we look at the top 20:
Death Guard x 1
Imperial Knights x 3
Astra Militarum x 3
Custodes x 1
Thousand Sons x 2
Eldar/Ynnari x 3
Dark Angels x 1
T’au x 2
Drukhari x 1
Cult Mechanicus x 1
Tyranids x 2
Some great data to look over from the event in this sheet, here.
Fun facts:
Most common Primary Factions:
Imperial Knights: 14
T’au: 13
Astra Militarum: 11
Asuryani: 11
Tyranids: 10
Most common Secondary Detachments:
Astra militarum: 40
T’au: 20
Asuryani: 19
Drukhari: 15
Tyranids: 12
Most points earned per round:
Adeptus Sororitas: 27.34
Dark Angels: 26.50
Imperial Knights: 24.12
Renegade Knights: 24
Ynnari: 23.85
Highest win percentage:
Renegade Knights: 75%
Dark Angels: 75:
Genestealer Cults: 58.33%
Drukhari: 55.32%
Imperial Knights: 55.28%
T’au: 55.07%
found it for you. Taking IG as your primary detachment meant on average you didnt not make the top for Win % or most points per round
Automatically Appended Next Post:
Marmatag wrote: The problem is you can't look at the field and draw a conclusion in regards to balance. And, i've seen that data, it did not account for "percentage used," it just had if a faction was used, at all. You are flatly misrepresenting this data. I would suggest you provide what you have.
You have to filter out the players who lose all, or most of their games. There's a reason people look at the lists that finish in the top 20 or so. Because those lists performed well. I don't care if B. Billy Bumshoes brought a "fluffy" list and got stomped out. That doesn't mean anything in regards to game balance. And the bottom of any tournament has a ton of total crap lists, especially if it's a big event.
You are actively arguing that we should induce survivorship bias into our statistical analysis
Asmodios wrote: A. You would need to make the strategems of equal power for both army A and B for this to become valid.
Yes you would and should! The idea that some armies have better stratagems by design rather than accident has absolutely no evidence. Sure, some armies have better stratagems, but some armies have better units too. It is not because GW intentionally tried to make it so, they just failed at balancing.
Also not sure how you can not agree that soup is inherently more powerful....... look at the number of soup armies in the top of tournaments and the number of mono guard...... there's a reason why all the top players for all the top factions are bringing soup and not mono guard lists. Take one look at nova and explain to me how soup isn't the issue 10/10 of the top lists are soup
"Yes, there is some extreme soup builds that are better than most (or even all) monobuilds, but I'd argue that most soup combinations are actually weaker than mono guard. If you punish soup in general while leaving guard as it is, then it just widens that gap. "
What part of this you didn't understand? Tournaments are representative of the absolutely cutting edge builds, and sure those seem to be some specific soup builds (mainly due poor balancing between codices.) It doesn't change the fact that excluding those couple of anomalies, mono guard is stronger than most soup builds. Do you think Black Templars + Sisters is stronger than mono-Guard? Ad Mech + Inquisition stronger than mono-Guard? No, they aren't. Sure Salamanders + Guard is stronger than pure Salamanders, but that's because Guard is disproportionately strong, and again pure Guard list would still be stronger.
What you are arguing is that mono guard is stronger then soup because of soup builds that we dont see and have no data on. Using that logic i could say that mono guard is awful because everyone we aren't looking at is spamming chimeras and sentinals. We have to work within the data data provided or we are simple just stating opinion with nothing to back up our claims
Nothing interesting about that top 10. It's exactly expected. 3 AM/ 2 Knights. Every single list including knights and AM some with BA - some with Custodes. A few Ynnari. A few Choas soupers. This is exactly expected. Also - top 10 has 3 AM primary...yet - you are trying to make a case that it hurts your win % to do this? Are you arguing in good faith here seriously?
Notice that primary guard did not make the top spots for win percentage or points earned per round. That primary guard preformed bellow mono armies like Tau. primary guard is stronger then mono guard and still finished bellow must take mono armies. This kind of evidence supports my previous claims that wile mono guard is good it is actually not supperior to armies like tau that strangely dont have 3 threads a day about how they need a nerf.
How are you determining Primary faction?
By points per faction?
By faction listed on BCP?
Number of detachments?
Or by largest detachment by points?
As all of the above give very diffrent impressions from the same data.
I've also seen peope list SoB as most points per faction but yiu look into it and its an msu battalion of sisters in like 5 lists none of which make the top 20? Are they really OP? No the data has just been manipulated incorrectly to make them.look OP.
Marmatag wrote: The problem is you can't look at the field and draw a conclusion in regards to balance. And, i've seen that data, it did not account for "percentage used," it just had if a faction was used, at all. You are flatly misrepresenting this data. I would suggest you provide what you have.
You have to filter out the players who lose all, or most of their games. There's a reason people look at the lists that finish in the top 20 or so. Because those lists performed well. I don't care if B. Billy Bumshoes brought a "fluffy" list and got stomped out. That doesn't mean anything in regards to game balance. And the bottom of any tournament has a ton of total crap lists, especially if it's a big event.
You are actively arguing that we should induce survivorship bias into our statistical analysis
You have absolutely no idea what you're talking about. Unless you're prepared to argue that the quality of lists and players is the same regardless of win/loss, you have no point whatsoever. My friend recently went 0-3 with Guard + Knights at an RTT. He forgot his models had invulnerable saves. As you can imagine, he's super casual and just enjoys the game. Should he count for the purposes of determining game balance?
You should actually go to a tournament. The bigger the tournament, the more casual players are there. They get weeded out early. The honest truth is if you are a competitive player and you consider yourself skilled, you should be 3-0 (at least 2-1) going into day 2 in any big event. It's unlikely you'll draw a nightmare matchup day 1 with a top 10 player.
You actually have no idea how statistics work or what survivorship bias is do you
You want to remove all the IG lists that do poorly and then compare the win % and points earned per round...... Seriously go plug survivorship bias into google do a bit of reading and just see how silly this comment is
No one audits it, really. As long as it looks close you can call your faction whatever you want. People declare their primary faction when they enter their list and register in BCP. In bigger tournaments where they award best in faction people keep a closer eye. But i could take my Tyranids to an RTT and play as Grey Knights.
And armies are judged by their ceilings, not their floors. This thread is beyond the pale.
Galas wrote: I remember when Marmatag was a nooby, all shiney and full of joy. How playing the unbalanced mess of Warhammer40k in a ultra-competitive meta has changed you
Bharring wrote: He'll actually *change his mind* in the presence of sufficient/sufficiently defended information. Heretic.
Xenomancers wrote: If that is true (i'd like to see the source) but what does that mean if the winning undefeated list are sporting IG brigades as their primary and often include more IG than the minimum? It clearly can't mean that putting more IG in your list is bad for your win rate.
First time using chess clocks at a gt/major on day 2 and despite leaning towards them not being the definitive answer before the event, I’ll admit I’m now firmly a supporter. Even in my last game when we had an issue with our clock working it was easy to call the judge over for a fix and it was flawless from there. Biggest benefit to them was that it keeps both players much more involved in the game imo. Doing so immediately gets the game moving and keeps it on track for a natural conclusion. None of my games were called to time. Round one: natural conclusion turn 4 – win
Round two: natural conclusion turn 3 – loss
Round three: natural conclusion turn 5 – win
Round four: natural conclusion turn 6 – win
Round five: natural conclusion turn 5 – loss
Round six: natural conclusion turn 6 – win
3 hour rounds were perfect. Holy crap does that extra few minutes help complete everything. In the past I’ve almost always felt rushed to complete my game, log the results and then run off to the next game with a stress filled running late feeling. Not the case with these round times. Adding 45 total minutes to the day was completely worth it imo.
So, we’ve got in the top 10:
Death Guard x 1
Imperial Knights x 2
Astra Militarum x 3
Custodes x 1
Thousand Sons x 1
Eldar/Ynnari x 2
Pretty interesting top 10 and a nice spread. It gets even more diverse when we look at the top 20:
Death Guard x 1
Imperial Knights x 3
Astra Militarum x 3
Custodes x 1
Thousand Sons x 2
Eldar/Ynnari x 3
Dark Angels x 1
T’au x 2
Drukhari x 1
Cult Mechanicus x 1
Tyranids x 2
Some great data to look over from the event in this sheet, here.
Fun facts:
Most common Primary Factions:
Imperial Knights: 14
T’au: 13
Astra Militarum: 11
Asuryani: 11
Tyranids: 10
Most common Secondary Detachments:
Astra militarum: 40
T’au: 20
Asuryani: 19
Drukhari: 15
Tyranids: 12
Most points earned per round:
Adeptus Sororitas: 27.34
Dark Angels: 26.50
Imperial Knights: 24.12
Renegade Knights: 24
Ynnari: 23.85
Highest win percentage:
Renegade Knights: 75%
Dark Angels: 75:
Genestealer Cults: 58.33%
Drukhari: 55.32%
Imperial Knights: 55.28%
T’au: 55.07%
found it for you. Taking IG as your primary detachment meant on average you didnt not make the top for Win % or most points per round
Automatically Appended Next Post:
Marmatag wrote: The problem is you can't look at the field and draw a conclusion in regards to balance. And, i've seen that data, it did not account for "percentage used," it just had if a faction was used, at all. You are flatly misrepresenting this data. I would suggest you provide what you have.
You have to filter out the players who lose all, or most of their games. There's a reason people look at the lists that finish in the top 20 or so. Because those lists performed well. I don't care if B. Billy Bumshoes brought a "fluffy" list and got stomped out. That doesn't mean anything in regards to game balance. And the bottom of any tournament has a ton of total crap lists, especially if it's a big event.
You are actively arguing that we should induce survivorship bias into our statistical analysis
Asmodios wrote: A. You would need to make the strategems of equal power for both army A and B for this to become valid.
Yes you would and should! The idea that some armies have better stratagems by design rather than accident has absolutely no evidence. Sure, some armies have better stratagems, but some armies have better units too. It is not because GW intentionally tried to make it so, they just failed at balancing.
Also not sure how you can not agree that soup is inherently more powerful....... look at the number of soup armies in the top of tournaments and the number of mono guard...... there's a reason why all the top players for all the top factions are bringing soup and not mono guard lists. Take one look at nova and explain to me how soup isn't the issue 10/10 of the top lists are soup
"Yes, there is some extreme soup builds that are better than most (or even all) monobuilds, but I'd argue that most soup combinations are actually weaker than mono guard. If you punish soup in general while leaving guard as it is, then it just widens that gap. "
What part of this you didn't understand? Tournaments are representative of the absolutely cutting edge builds, and sure those seem to be some specific soup builds (mainly due poor balancing between codices.) It doesn't change the fact that excluding those couple of anomalies, mono guard is stronger than most soup builds. Do you think Black Templars + Sisters is stronger than mono-Guard? Ad Mech + Inquisition stronger than mono-Guard? No, they aren't. Sure Salamanders + Guard is stronger than pure Salamanders, but that's because Guard is disproportionately strong, and again pure Guard list would still be stronger.
What you are arguing is that mono guard is stronger then soup because of soup builds that we dont see and have no data on. Using that logic i could say that mono guard is awful because everyone we aren't looking at is spamming chimeras and sentinals. We have to work within the data data provided or we are simple just stating opinion with nothing to back up our claims
Nothing interesting about that top 10. It's exactly expected. 3 AM/ 2 Knights. Every single list including knights and AM some with BA - some with Custodes. A few Ynnari. A few Choas soupers. This is exactly expected. Also - top 10 has 3 AM primary...yet - you are trying to make a case that it hurts your win % to do this? Are you arguing in good faith here seriously?
Notice that primary guard did not make the top spots for win percentage or points earned per round. That primary guard preformed bellow mono armies like Tau. primary guard is stronger then mono guard and still finished bellow must take mono armies. This kind of evidence supports my previous claims that wile mono guard is good it is actually not supperior to armies like tau that strangely dont have 3 threads a day about how they need a nerf.
How are you determining Primary faction?
By points per faction?
By faction listed on BCP?
Number of detachments?
Or by largest detachment by points?
As all of the above give very diffrent impressions from the same data.
I've also seen peope list SoB as most points per faction but yiu look into it and its an msu battalion of sisters in like 5 lists none of which make the top 20? Are they really OP? No the data has just been manipulated incorrectly to make them.look OP.
For this data the defined primary faction as the faction the made up the largest % of your army so if Faction A was 999 points and faction B was 1001 points faction B would be your primary faction. So including guard as your primary faction in this tournament lowered your win % and points per round compared to those that took things like knights as their primary faction
Automatically Appended Next Post:
Marmatag wrote: No one audits it, really. As long as it looks close you can call your faction whatever you want. People declare their primary faction when they enter their list and register in BCP. In bigger tournaments where they award best in faction people keep a closer eye. But i could take my Tyranids to an RTT and play as Grey Knights.
And armies are judged by their ceilings, not their floors. This thread is beyond the pale.
1. The primary faction was done by the army that made up the largest % of your army...... so you have no point there
2. What your asking to do is to remove the left side of the bell curve for a statistic and then expect your results to not be off. Its a real term in a statistical analysis called survivorship bias and it causes you to have inaccurate results. You are literally arguing that we should skew the data in hopes of getting the result we want.
This message was edited 1 time. Last update was at 2018/10/04 19:42:23
Except your analysis doesn't take into account the quality of the list. Some lists are bad. Some lists are good. Your analysis conflates lists by Keyword, when in reality armies like Guard, Tau, etc, have a multitude of different ways they can be played. Stop being obtuse.
And again, do you think BCP magically scans the screenshot of a napkin with coffee stains on it where the list was written? It would have to be a player reporting a misrepresented faction for it to be caught. I know a guy who accidentally misreported his faction and it wasn't caught until way later after the event.
Galas wrote: I remember when Marmatag was a nooby, all shiney and full of joy. How playing the unbalanced mess of Warhammer40k in a ultra-competitive meta has changed you
Bharring wrote: He'll actually *change his mind* in the presence of sufficient/sufficiently defended information. Heretic.
Just looking at the results and not accounting for actual player skill and list composition is also skewing the results.
Sure if all generals were the same and IG soup meant the same thing then your statistical analysis would be awesome. But that doesn't happen in tournaments. Frank wants to play his IG tank battalion and refuses to bring anything else to the tournament. That list counts among the IG power bell-curve but does nothing to address how powerful/cutting edge/op army lists are doing with good generals (which is what we are trying to control for when we talk of a factions power, right?)
Taking your results from the top 20% or so of tourney results usually filters out the casual guys, the people there just for a good time or people that drop out after the first day because they've gotten their yearly quota of butt kicking.
I'm not sure how this isn't obvious to anyone whose been to a tournament.
Marmatag wrote: Except your analysis doesn't take into account the quality of the list. Some lists are bad. Some lists are good. Your analysis conflates lists by Keyword, when in reality armies like Guard, Tau, etc, have a multitude of different ways they can be played. Stop being obtuse.
And again, do you think BCP magically scans the screenshot of a napkin with coffee stains on it where the list was written? It would have to be a player reporting a misrepresented faction for it to be caught. I know a guy who accidentally misreported his faction and it wasn't caught until way later after the event.
1. They covered how the data was gathered in that weeks FLG podcast. They took every list plugged it into excel and ran that stats
2. Once again you skew the data by removing the left side of the bell curve. For instance only look at the undefeated DG player from BAO... remove all other DG players..... ZOMG DG has 100% win rate they cannot be beaten nerf blightlord terminators now!!!! skewing the data on purpose gives you a poor view of win percentages. There is a reason why no statistical analysis does this. I would have failed every stats class i took in college had i run regressions this way
bananathug wrote: Just looking at the results and not accounting for actual player skill and list composition is also skewing the results.
Sure if all generals were the same and IG soup meant the same thing then your statistical analysis would be awesome. But that doesn't happen in tournaments. Frank wants to play his IG tank battalion and refuses to bring anything else to the tournament. That list counts among the IG power bell-curve but does nothing to address how powerful/cutting edge/op army lists are doing with good generals (which is what we are trying to control for when we talk of a factions power, right?)
Taking your results from the top 20% or so of tourney results usually filters out the casual guys, the people there just for a good time or people that drop out after the first day because they've gotten their yearly quota of butt kicking.
I'm not sure how this isn't obvious to anyone whose been to a tournament.
Thank you. This thread is nuts.
"MARY J. feth FACE WENT 0-5 WITH HER FIRE DRAGON THEMED ELDAR LIST WITH 0 PSYKERS. WHO KNEW RUNNING TRIPLE BIEL TAN WRAITHKNIGHTS WOULDNT WORK. BUFF CWE."
This message was edited 3 times. Last update was at 2018/10/04 19:58:53
Galas wrote: I remember when Marmatag was a nooby, all shiney and full of joy. How playing the unbalanced mess of Warhammer40k in a ultra-competitive meta has changed you
Bharring wrote: He'll actually *change his mind* in the presence of sufficient/sufficiently defended information. Heretic.
bananathug wrote: Just looking at the results and not accounting for actual player skill and list composition is also skewing the results.
Sure if all generals were the same and IG soup meant the same thing then your statistical analysis would be awesome. But that doesn't happen in tournaments. Frank wants to play his IG tank battalion and refuses to bring anything else to the tournament. That list counts among the IG power bell-curve but does nothing to address how powerful/cutting edge/op army lists are doing with good generals (which is what we are trying to control for when we talk of a factions power, right?)
Taking your results from the top 20% or so of tourney results usually filters out the casual guys, the people there just for a good time or people that drop out after the first day because they've gotten their yearly quota of butt kicking.
I'm not sure how this isn't obvious to anyone whose been to a tournament.
You cannot do this with a data pool this size. What you want to do is hold variables constant (which is perfectly legitimate). The issue is there is no data that allows you to do this effectively.
You would either need
.A thousands of repetitions of same said tournament so you could get an actual representation of "skill".
.B Control stats where x "skilled player" played every different army so you could analyze exactly how high he could push any given army
The fact is that cutting out half the data at a tournament would greatly skew the data. Especially when the variation between armies played is so high and there's a minimum number of rounds
bananathug wrote: Just looking at the results and not accounting for actual player skill and list composition is also skewing the results.
Sure if all generals were the same and IG soup meant the same thing then your statistical analysis would be awesome. But that doesn't happen in tournaments. Frank wants to play his IG tank battalion and refuses to bring anything else to the tournament. That list counts among the IG power bell-curve but does nothing to address how powerful/cutting edge/op army lists are doing with good generals (which is what we are trying to control for when we talk of a factions power, right?)
Taking your results from the top 20% or so of tourney results usually filters out the casual guys, the people there just for a good time or people that drop out after the first day because they've gotten their yearly quota of butt kicking.
I'm not sure how this isn't obvious to anyone whose been to a tournament.
Thank you. This thread is nuts.
"MARY J. feth FACE WENT 0-5 WITH HER FIRE DRAGON THEMED ELDAR LIST WITH 0 PSYKERS. WHO KNEW RUNNING TRIPLE BIEL TAN WRAITHKNIGHTS WOULDNT WORK. BUFF CWE."
>We want to see if men or women are better drivers
>Remove the bottom 90% of bad drivers across the board
>expect good results
This message was edited 1 time. Last update was at 2018/10/04 20:02:31
Xenomancers wrote: If that is true (i'd like to see the source) but what does that mean if the winning undefeated list are sporting IG brigades as their primary and often include more IG than the minimum? It clearly can't mean that putting more IG in your list is bad for your win rate.
First time using chess clocks at a gt/major on day 2 and despite leaning towards them not being the definitive answer before the event, I’ll admit I’m now firmly a supporter. Even in my last game when we had an issue with our clock working it was easy to call the judge over for a fix and it was flawless from there. Biggest benefit to them was that it keeps both players much more involved in the game imo. Doing so immediately gets the game moving and keeps it on track for a natural conclusion. None of my games were called to time. Round one: natural conclusion turn 4 – win
Round two: natural conclusion turn 3 – loss
Round three: natural conclusion turn 5 – win
Round four: natural conclusion turn 6 – win
Round five: natural conclusion turn 5 – loss
Round six: natural conclusion turn 6 – win
3 hour rounds were perfect. Holy crap does that extra few minutes help complete everything. In the past I’ve almost always felt rushed to complete my game, log the results and then run off to the next game with a stress filled running late feeling. Not the case with these round times. Adding 45 total minutes to the day was completely worth it imo.
So, we’ve got in the top 10:
Death Guard x 1
Imperial Knights x 2
Astra Militarum x 3
Custodes x 1
Thousand Sons x 1
Eldar/Ynnari x 2
Pretty interesting top 10 and a nice spread. It gets even more diverse when we look at the top 20:
Death Guard x 1
Imperial Knights x 3
Astra Militarum x 3
Custodes x 1
Thousand Sons x 2
Eldar/Ynnari x 3
Dark Angels x 1
T’au x 2
Drukhari x 1
Cult Mechanicus x 1
Tyranids x 2
Some great data to look over from the event in this sheet, here.
Fun facts:
Most common Primary Factions:
Imperial Knights: 14
T’au: 13
Astra Militarum: 11
Asuryani: 11
Tyranids: 10
Most common Secondary Detachments:
Astra militarum: 40
T’au: 20
Asuryani: 19
Drukhari: 15
Tyranids: 12
Most points earned per round:
Adeptus Sororitas: 27.34
Dark Angels: 26.50
Imperial Knights: 24.12
Renegade Knights: 24
Ynnari: 23.85
Highest win percentage:
Renegade Knights: 75%
Dark Angels: 75:
Genestealer Cults: 58.33%
Drukhari: 55.32%
Imperial Knights: 55.28%
T’au: 55.07%
found it for you. Taking IG as your primary detachment meant on average you didnt not make the top for Win % or most points per round
Automatically Appended Next Post:
Marmatag wrote: The problem is you can't look at the field and draw a conclusion in regards to balance. And, i've seen that data, it did not account for "percentage used," it just had if a faction was used, at all. You are flatly misrepresenting this data. I would suggest you provide what you have.
You have to filter out the players who lose all, or most of their games. There's a reason people look at the lists that finish in the top 20 or so. Because those lists performed well. I don't care if B. Billy Bumshoes brought a "fluffy" list and got stomped out. That doesn't mean anything in regards to game balance. And the bottom of any tournament has a ton of total crap lists, especially if it's a big event.
You are actively arguing that we should induce survivorship bias into our statistical analysis
Asmodios wrote: A. You would need to make the strategems of equal power for both army A and B for this to become valid.
Yes you would and should! The idea that some armies have better stratagems by design rather than accident has absolutely no evidence. Sure, some armies have better stratagems, but some armies have better units too. It is not because GW intentionally tried to make it so, they just failed at balancing.
Also not sure how you can not agree that soup is inherently more powerful....... look at the number of soup armies in the top of tournaments and the number of mono guard...... there's a reason why all the top players for all the top factions are bringing soup and not mono guard lists. Take one look at nova and explain to me how soup isn't the issue 10/10 of the top lists are soup
"Yes, there is some extreme soup builds that are better than most (or even all) monobuilds, but I'd argue that most soup combinations are actually weaker than mono guard. If you punish soup in general while leaving guard as it is, then it just widens that gap. "
What part of this you didn't understand? Tournaments are representative of the absolutely cutting edge builds, and sure those seem to be some specific soup builds (mainly due poor balancing between codices.) It doesn't change the fact that excluding those couple of anomalies, mono guard is stronger than most soup builds. Do you think Black Templars + Sisters is stronger than mono-Guard? Ad Mech + Inquisition stronger than mono-Guard? No, they aren't. Sure Salamanders + Guard is stronger than pure Salamanders, but that's because Guard is disproportionately strong, and again pure Guard list would still be stronger.
What you are arguing is that mono guard is stronger then soup because of soup builds that we dont see and have no data on. Using that logic i could say that mono guard is awful because everyone we aren't looking at is spamming chimeras and sentinals. We have to work within the data data provided or we are simple just stating opinion with nothing to back up our claims
Nothing interesting about that top 10. It's exactly expected. 3 AM/ 2 Knights. Every single list including knights and AM some with BA - some with Custodes. A few Ynnari. A few Choas soupers. This is exactly expected. Also - top 10 has 3 AM primary...yet - you are trying to make a case that it hurts your win % to do this? Are you arguing in good faith here seriously?
Notice that primary guard did not make the top spots for win percentage or points earned per round. That primary guard preformed bellow mono armies like Tau. primary guard is stronger then mono guard and still finished bellow must take mono armies. This kind of evidence supports my previous claims that wile mono guard is good it is actually not supperior to armies like tau that strangely dont have 3 threads a day about how they need a nerf.
Tau actually have a lot more threads about how much they suck. LOL. I'm not going to argue that tau isn't strong as a mono force. I'd rate them in the top 5 for mono army. I am considering taking Tau to LVO just to see how well I can do. Though - I would really like to have a chance to win it all (even if it is slim) so I will probably be bringing some sort of knight variation depending on what chapter approved does.
Anyways. Tau are certainly a strong army. I win a lot more than I lose with them - and it's usually not close. I felt so bad playing against my good friends mono Custodes army - the only units I lost all game. A single ghost keel (which I kind of let him kill) and 3 fire warriors and 3 drones.
Meanwhile I destroyed a squad of bikes and his super dread (I forget it's name but it's basically a leviathan) and a unit of guards all at -1 to hit the first turn (and I didn't roll very good ether). Basically GG right there.
Marmatag wrote: Except your analysis doesn't take into account the quality of the list. Some lists are bad. Some lists are good. Your analysis conflates lists by Keyword, when in reality armies like Guard, Tau, etc, have a multitude of different ways they can be played. Stop being obtuse.
And again, do you think BCP magically scans the screenshot of a napkin with coffee stains on it where the list was written? It would have to be a player reporting a misrepresented faction for it to be caught. I know a guy who accidentally misreported his faction and it wasn't caught until way later after the event.
1. They covered how the data was gathered in that weeks FLG podcast. They took every list plugged it into excel and ran that stats
2. Once again you skew the data by removing the left side of the bell curve. For instance only look at the undefeated DG player from BAO... remove all other DG players..... ZOMG DG has 100% win rate they cannot be beaten nerf blightlord terminators now!!!! skewing the data on purpose gives you a poor view of win percentages. There is a reason why no statistical analysis does this. I would have failed every stats class i took in college had i run regressions this way
Most statistical analysis removes outliers. That gives you the best idea of the averages. We don't to know whos average - we want to know whos best. So you should remove the bottom 50% I say only look at the top 20%. The trash armies at the bottom offer nothing to the question of what is best.
This message was edited 3 times. Last update was at 2018/10/04 20:14:05
If we fail to anticipate the unforeseen or expect the unexpected in a universe of infinite possibilities, we may find ourselves at the mercy of anyone or anything that cannot be programmed, categorized or easily referenced.
- Fox Mulder
In the beginning of 8th, I understand why ITC dropped the largest detachment in points was your primary faction and let you pick what you wanted, since everyone had interdetachment soup. But now that all detachments are mono factions, they could easily go back to largest detachment in points is primary faction as they did in 7th. It certainly makes their rankings more honest.
Xenomancers wrote: If that is true (i'd like to see the source) but what does that mean if the winning undefeated list are sporting IG brigades as their primary and often include more IG than the minimum? It clearly can't mean that putting more IG in your list is bad for your win rate.
First time using chess clocks at a gt/major on day 2 and despite leaning towards them not being the definitive answer before the event, I’ll admit I’m now firmly a supporter. Even in my last game when we had an issue with our clock working it was easy to call the judge over for a fix and it was flawless from there. Biggest benefit to them was that it keeps both players much more involved in the game imo. Doing so immediately gets the game moving and keeps it on track for a natural conclusion. None of my games were called to time. Round one: natural conclusion turn 4 – win
Round two: natural conclusion turn 3 – loss
Round three: natural conclusion turn 5 – win
Round four: natural conclusion turn 6 – win
Round five: natural conclusion turn 5 – loss
Round six: natural conclusion turn 6 – win
3 hour rounds were perfect. Holy crap does that extra few minutes help complete everything. In the past I’ve almost always felt rushed to complete my game, log the results and then run off to the next game with a stress filled running late feeling. Not the case with these round times. Adding 45 total minutes to the day was completely worth it imo.
So, we’ve got in the top 10:
Death Guard x 1
Imperial Knights x 2
Astra Militarum x 3
Custodes x 1
Thousand Sons x 1
Eldar/Ynnari x 2
Pretty interesting top 10 and a nice spread. It gets even more diverse when we look at the top 20:
Death Guard x 1
Imperial Knights x 3
Astra Militarum x 3
Custodes x 1
Thousand Sons x 2
Eldar/Ynnari x 3
Dark Angels x 1
T’au x 2
Drukhari x 1
Cult Mechanicus x 1
Tyranids x 2
Some great data to look over from the event in this sheet, here.
Fun facts:
Most common Primary Factions:
Imperial Knights: 14
T’au: 13
Astra Militarum: 11
Asuryani: 11
Tyranids: 10
Most common Secondary Detachments:
Astra militarum: 40
T’au: 20
Asuryani: 19
Drukhari: 15
Tyranids: 12
Most points earned per round:
Adeptus Sororitas: 27.34
Dark Angels: 26.50
Imperial Knights: 24.12
Renegade Knights: 24
Ynnari: 23.85
Highest win percentage:
Renegade Knights: 75%
Dark Angels: 75:
Genestealer Cults: 58.33%
Drukhari: 55.32%
Imperial Knights: 55.28%
T’au: 55.07%
found it for you. Taking IG as your primary detachment meant on average you didnt not make the top for Win % or most points per round
Automatically Appended Next Post:
Marmatag wrote: The problem is you can't look at the field and draw a conclusion in regards to balance. And, i've seen that data, it did not account for "percentage used," it just had if a faction was used, at all. You are flatly misrepresenting this data. I would suggest you provide what you have.
You have to filter out the players who lose all, or most of their games. There's a reason people look at the lists that finish in the top 20 or so. Because those lists performed well. I don't care if B. Billy Bumshoes brought a "fluffy" list and got stomped out. That doesn't mean anything in regards to game balance. And the bottom of any tournament has a ton of total crap lists, especially if it's a big event.
You are actively arguing that we should induce survivorship bias into our statistical analysis
Asmodios wrote: A. You would need to make the strategems of equal power for both army A and B for this to become valid.
Yes you would and should! The idea that some armies have better stratagems by design rather than accident has absolutely no evidence. Sure, some armies have better stratagems, but some armies have better units too. It is not because GW intentionally tried to make it so, they just failed at balancing.
Also not sure how you can not agree that soup is inherently more powerful....... look at the number of soup armies in the top of tournaments and the number of mono guard...... there's a reason why all the top players for all the top factions are bringing soup and not mono guard lists. Take one look at nova and explain to me how soup isn't the issue 10/10 of the top lists are soup
"Yes, there is some extreme soup builds that are better than most (or even all) monobuilds, but I'd argue that most soup combinations are actually weaker than mono guard. If you punish soup in general while leaving guard as it is, then it just widens that gap. "
What part of this you didn't understand? Tournaments are representative of the absolutely cutting edge builds, and sure those seem to be some specific soup builds (mainly due poor balancing between codices.) It doesn't change the fact that excluding those couple of anomalies, mono guard is stronger than most soup builds. Do you think Black Templars + Sisters is stronger than mono-Guard? Ad Mech + Inquisition stronger than mono-Guard? No, they aren't. Sure Salamanders + Guard is stronger than pure Salamanders, but that's because Guard is disproportionately strong, and again pure Guard list would still be stronger.
What you are arguing is that mono guard is stronger then soup because of soup builds that we dont see and have no data on. Using that logic i could say that mono guard is awful because everyone we aren't looking at is spamming chimeras and sentinals. We have to work within the data data provided or we are simple just stating opinion with nothing to back up our claims
Nothing interesting about that top 10. It's exactly expected. 3 AM/ 2 Knights. Every single list including knights and AM some with BA - some with Custodes. A few Ynnari. A few Choas soupers. This is exactly expected. Also - top 10 has 3 AM primary...yet - you are trying to make a case that it hurts your win % to do this? Are you arguing in good faith here seriously?
Notice that primary guard did not make the top spots for win percentage or points earned per round. That primary guard preformed bellow mono armies like Tau. primary guard is stronger then mono guard and still finished bellow must take mono armies. This kind of evidence supports my previous claims that wile mono guard is good it is actually not supperior to armies like tau that strangely dont have 3 threads a day about how they need a nerf.
Tau actually have a lot more threads about how much they suck. LOL. I'm not going to argue that tau isn't strong as a mono force. I'd rate them in the top 5 for mono army. I am considering taking Tau to LVO just to see how well I can do. Though - I would really like to have a chance to win it all (even if it is slim) so I will probably be bringing some sort of knight variation depending on what chapter approved does.
Anyways. Tau are certainly a strong army. I win a lot more than I lose with them - and it's usually not close. I felt so bad playing against my good friends mono Custodes army - the only units I lost all game. A single ghost keel (which I kind of let him kill) and 3 fire warriors and 3 drones.
Meanwhile I destroyed a squad of bikes and his super dread (I forget it's name but it's basically a leviathan) and a unit of guards all at -1 to hit the first turn (and I didn't roll very good ether). Basically GG right there.
Marmatag wrote: Except your analysis doesn't take into account the quality of the list. Some lists are bad. Some lists are good. Your analysis conflates lists by Keyword, when in reality armies like Guard, Tau, etc, have a multitude of different ways they can be played. Stop being obtuse.
And again, do you think BCP magically scans the screenshot of a napkin with coffee stains on it where the list was written? It would have to be a player reporting a misrepresented faction for it to be caught. I know a guy who accidentally misreported his faction and it wasn't caught until way later after the event.
1. They covered how the data was gathered in that weeks FLG podcast. They took every list plugged it into excel and ran that stats
2. Once again you skew the data by removing the left side of the bell curve. For instance only look at the undefeated DG player from BAO... remove all other DG players..... ZOMG DG has 100% win rate they cannot be beaten nerf blightlord terminators now!!!! skewing the data on purpose gives you a poor view of win percentages. There is a reason why no statistical analysis does this. I would have failed every stats class i took in college had i run regressions this way
Most statistical analysis removes outliers. That gives you the best idea of the averages. We don't to know whos average - we want to know whos best. So you should remove the bottom 50% I say only look at the top 20%. The trash armies at the bottom offer nothing to the question of what is best.
That's not "removing an outlier" and my post above shows why it would incredibly skew data in such a small sample size. For example, the DG would shoot up to a 100% win rate. To do it correctly too you would need to repeat the exact tournament with the same opponents over and over while having them play all available army so that way you could get a true "player skill". Even then under perfect conditions, you are changing the definition of "best army/ faction" To "best army/ faction while being played by top 10%" meaning that the data would really only be useful when taken by top 10-20%. You also run into the issue of the top 10-20% not even bringing certain factions so losing data on where those fall.
At the end of the day skewing this sample in any of those ways is going to give you a worse picture of the data.
Xenomancers wrote: If that is true (i'd like to see the source) but what does that mean if the winning undefeated list are sporting IG brigades as their primary and often include more IG than the minimum? It clearly can't mean that putting more IG in your list is bad for your win rate.
First time using chess clocks at a gt/major on day 2 and despite leaning towards them not being the definitive answer before the event, I’ll admit I’m now firmly a supporter. Even in my last game when we had an issue with our clock working it was easy to call the judge over for a fix and it was flawless from there. Biggest benefit to them was that it keeps both players much more involved in the game imo. Doing so immediately gets the game moving and keeps it on track for a natural conclusion. None of my games were called to time. Round one: natural conclusion turn 4 – win
Round two: natural conclusion turn 3 – loss
Round three: natural conclusion turn 5 – win
Round four: natural conclusion turn 6 – win
Round five: natural conclusion turn 5 – loss
Round six: natural conclusion turn 6 – win
3 hour rounds were perfect. Holy crap does that extra few minutes help complete everything. In the past I’ve almost always felt rushed to complete my game, log the results and then run off to the next game with a stress filled running late feeling. Not the case with these round times. Adding 45 total minutes to the day was completely worth it imo.
So, we’ve got in the top 10:
Death Guard x 1
Imperial Knights x 2
Astra Militarum x 3
Custodes x 1
Thousand Sons x 1
Eldar/Ynnari x 2
Pretty interesting top 10 and a nice spread. It gets even more diverse when we look at the top 20:
Death Guard x 1
Imperial Knights x 3
Astra Militarum x 3
Custodes x 1
Thousand Sons x 2
Eldar/Ynnari x 3
Dark Angels x 1
T’au x 2
Drukhari x 1
Cult Mechanicus x 1
Tyranids x 2
Some great data to look over from the event in this sheet, here.
Fun facts:
Most common Primary Factions:
Imperial Knights: 14
T’au: 13
Astra Militarum: 11
Asuryani: 11
Tyranids: 10
Most common Secondary Detachments:
Astra militarum: 40
T’au: 20
Asuryani: 19
Drukhari: 15
Tyranids: 12
Most points earned per round:
Adeptus Sororitas: 27.34
Dark Angels: 26.50
Imperial Knights: 24.12
Renegade Knights: 24
Ynnari: 23.85
Highest win percentage:
Renegade Knights: 75%
Dark Angels: 75:
Genestealer Cults: 58.33%
Drukhari: 55.32%
Imperial Knights: 55.28%
T’au: 55.07%
found it for you. Taking IG as your primary detachment meant on average you didnt not make the top for Win % or most points per round
Automatically Appended Next Post:
Marmatag wrote: The problem is you can't look at the field and draw a conclusion in regards to balance. And, i've seen that data, it did not account for "percentage used," it just had if a faction was used, at all. You are flatly misrepresenting this data. I would suggest you provide what you have.
You have to filter out the players who lose all, or most of their games. There's a reason people look at the lists that finish in the top 20 or so. Because those lists performed well. I don't care if B. Billy Bumshoes brought a "fluffy" list and got stomped out. That doesn't mean anything in regards to game balance. And the bottom of any tournament has a ton of total crap lists, especially if it's a big event.
You are actively arguing that we should induce survivorship bias into our statistical analysis
Asmodios wrote: A. You would need to make the strategems of equal power for both army A and B for this to become valid.
Yes you would and should! The idea that some armies have better stratagems by design rather than accident has absolutely no evidence. Sure, some armies have better stratagems, but some armies have better units too. It is not because GW intentionally tried to make it so, they just failed at balancing.
Also not sure how you can not agree that soup is inherently more powerful....... look at the number of soup armies in the top of tournaments and the number of mono guard...... there's a reason why all the top players for all the top factions are bringing soup and not mono guard lists. Take one look at nova and explain to me how soup isn't the issue 10/10 of the top lists are soup
"Yes, there is some extreme soup builds that are better than most (or even all) monobuilds, but I'd argue that most soup combinations are actually weaker than mono guard. If you punish soup in general while leaving guard as it is, then it just widens that gap. "
What part of this you didn't understand? Tournaments are representative of the absolutely cutting edge builds, and sure those seem to be some specific soup builds (mainly due poor balancing between codices.) It doesn't change the fact that excluding those couple of anomalies, mono guard is stronger than most soup builds. Do you think Black Templars + Sisters is stronger than mono-Guard? Ad Mech + Inquisition stronger than mono-Guard? No, they aren't. Sure Salamanders + Guard is stronger than pure Salamanders, but that's because Guard is disproportionately strong, and again pure Guard list would still be stronger.
What you are arguing is that mono guard is stronger then soup because of soup builds that we dont see and have no data on. Using that logic i could say that mono guard is awful because everyone we aren't looking at is spamming chimeras and sentinals. We have to work within the data data provided or we are simple just stating opinion with nothing to back up our claims
Nothing interesting about that top 10. It's exactly expected. 3 AM/ 2 Knights. Every single list including knights and AM some with BA - some with Custodes. A few Ynnari. A few Choas soupers. This is exactly expected. Also - top 10 has 3 AM primary...yet - you are trying to make a case that it hurts your win % to do this? Are you arguing in good faith here seriously?
Notice that primary guard did not make the top spots for win percentage or points earned per round. That primary guard preformed bellow mono armies like Tau. primary guard is stronger then mono guard and still finished bellow must take mono armies. This kind of evidence supports my previous claims that wile mono guard is good it is actually not supperior to armies like tau that strangely dont have 3 threads a day about how they need a nerf.
Tau actually have a lot more threads about how much they suck. LOL. I'm not going to argue that tau isn't strong as a mono force. I'd rate them in the top 5 for mono army. I am considering taking Tau to LVO just to see how well I can do. Though - I would really like to have a chance to win it all (even if it is slim) so I will probably be bringing some sort of knight variation depending on what chapter approved does.
Anyways. Tau are certainly a strong army. I win a lot more than I lose with them - and it's usually not close. I felt so bad playing against my good friends mono Custodes army - the only units I lost all game. A single ghost keel (which I kind of let him kill) and 3 fire warriors and 3 drones.
Meanwhile I destroyed a squad of bikes and his super dread (I forget it's name but it's basically a leviathan) and a unit of guards all at -1 to hit the first turn (and I didn't roll very good ether). Basically GG right there.
Marmatag wrote: Except your analysis doesn't take into account the quality of the list. Some lists are bad. Some lists are good. Your analysis conflates lists by Keyword, when in reality armies like Guard, Tau, etc, have a multitude of different ways they can be played. Stop being obtuse.
And again, do you think BCP magically scans the screenshot of a napkin with coffee stains on it where the list was written? It would have to be a player reporting a misrepresented faction for it to be caught. I know a guy who accidentally misreported his faction and it wasn't caught until way later after the event.
1. They covered how the data was gathered in that weeks FLG podcast. They took every list plugged it into excel and ran that stats
2. Once again you skew the data by removing the left side of the bell curve. For instance only look at the undefeated DG player from BAO... remove all other DG players..... ZOMG DG has 100% win rate they cannot be beaten nerf blightlord terminators now!!!! skewing the data on purpose gives you a poor view of win percentages. There is a reason why no statistical analysis does this. I would have failed every stats class i took in college had i run regressions this way
Most statistical analysis removes outliers. That gives you the best idea of the averages. We don't to know whos average - we want to know whos best. So you should remove the bottom 50% I say only look at the top 20%. The trash armies at the bottom offer nothing to the question of what is best.
That's not "removing an outlier" and my post above shows why it would incredibly skew data in such a small sample size. For example, the DG would shoot up to a 100% win rate. To do it correctly too you would need to repeat the exact tournament with the same opponents over and over while having them play all available army so that way you could get a true "player skill". Even then under perfect conditions, you are changing the definition of "best army/ faction" To "best army/ faction while being played by top 10%" meaning that the data would really only be useful when taken by top 10-20%. You also run into the issue of the top 10-20% not even bringing certain factions so losing data on where those fall.
At the end of the day skewing this sample in any of those ways is going to give you a worse picture of the data.
[/spoiler]
Wouldn't it be better to cut the lowest part of the tournament results AND the top tournament results? I mean, i'm an engineer not a math expert, but that's what i would do.
Because we're looking at the potential of each army with the best configuration(s).
No one cares you can run sub-optimal lists. I'll openly acknowledge every codex can suck if you screw it up hard enough. But that's not the point, because the game isn't about racing to the bottom.
Galas wrote: I remember when Marmatag was a nooby, all shiney and full of joy. How playing the unbalanced mess of Warhammer40k in a ultra-competitive meta has changed you
Bharring wrote: He'll actually *change his mind* in the presence of sufficient/sufficiently defended information. Heretic.
Marmatag wrote: Because we're looking at the potential of each army with the best configuration(s).
No one cares you can run sub-optimal lists. I'll openly acknowledge every codex can suck if you screw it up hard enough. But that's not the point, because the game isn't about racing to the bottom.
Not everyone is concerned about the top 10% though.
Some of us are more concerned about what happens if we bring our favourite army to a pick up game at our local store.
Asmodios wrote: That's not "removing an outlier" and my post above shows why it would incredibly skew data in such a small sample size. For example, the DG would shoot up to a 100% win rate. To do it correctly too you would need to repeat the exact tournament with the same opponents over and over while having them play all available army so that way you could get a true "player skill". Even then under perfect conditions, you are changing the definition of "best army/ faction" To "best army/ faction while being played by top 10%" meaning that the data would really only be useful when taken by top 10-20%. You also run into the issue of the top 10-20% not even bringing certain factions so losing data on where those fall.
At the end of the day skewing this sample in any of those ways is going to give you a worse picture of the data.
You don't base everything on one tournament. That's one example of badly sampling the data. Flukes do happen.
We know there are good and bad units and better or worse factions and soups as a result. We know there are good and bad players.
Theoretically therefore good players, with a list you have worked out should be good, would be expected to place high.
Testing: Do they, at multiple events. Oh look yes.
Conclusion - these units are good.
This sort of "we can't really know anything" is reading like Eldar in 7th.
Marmatag wrote: Because we're looking at the potential of each army with the best configuration(s).
No one cares you can run sub-optimal lists. I'll openly acknowledge every codex can suck if you screw it up hard enough. But that's not the point, because the game isn't about racing to the bottom.
Not everyone is concerned about the top 10% though.
Some of us are more concerned about what happens if we bring our favourite army to a pick up game at our local store.
Then you should find that out by doing it, not by hunting for broken combos on the internet. Netlisting sucks especially in a small pond meta.
This message was edited 2 times. Last update was at 2018/10/04 21:19:20
Galas wrote: I remember when Marmatag was a nooby, all shiney and full of joy. How playing the unbalanced mess of Warhammer40k in a ultra-competitive meta has changed you
Bharring wrote: He'll actually *change his mind* in the presence of sufficient/sufficiently defended information. Heretic.
Xenomancers wrote: If that is true (i'd like to see the source) but what does that mean if the winning undefeated list are sporting IG brigades as their primary and often include more IG than the minimum? It clearly can't mean that putting more IG in your list is bad for your win rate.
First time using chess clocks at a gt/major on day 2 and despite leaning towards them not being the definitive answer before the event, I’ll admit I’m now firmly a supporter. Even in my last game when we had an issue with our clock working it was easy to call the judge over for a fix and it was flawless from there. Biggest benefit to them was that it keeps both players much more involved in the game imo. Doing so immediately gets the game moving and keeps it on track for a natural conclusion. None of my games were called to time. Round one: natural conclusion turn 4 – win
Round two: natural conclusion turn 3 – loss
Round three: natural conclusion turn 5 – win
Round four: natural conclusion turn 6 – win
Round five: natural conclusion turn 5 – loss
Round six: natural conclusion turn 6 – win
3 hour rounds were perfect. Holy crap does that extra few minutes help complete everything. In the past I’ve almost always felt rushed to complete my game, log the results and then run off to the next game with a stress filled running late feeling. Not the case with these round times. Adding 45 total minutes to the day was completely worth it imo.
So, we’ve got in the top 10:
Death Guard x 1
Imperial Knights x 2
Astra Militarum x 3
Custodes x 1
Thousand Sons x 1
Eldar/Ynnari x 2
Pretty interesting top 10 and a nice spread. It gets even more diverse when we look at the top 20:
Death Guard x 1
Imperial Knights x 3
Astra Militarum x 3
Custodes x 1
Thousand Sons x 2
Eldar/Ynnari x 3
Dark Angels x 1
T’au x 2
Drukhari x 1
Cult Mechanicus x 1
Tyranids x 2
Some great data to look over from the event in this sheet, here.
Fun facts:
Most common Primary Factions:
Imperial Knights: 14
T’au: 13
Astra Militarum: 11
Asuryani: 11
Tyranids: 10
Most common Secondary Detachments:
Astra militarum: 40
T’au: 20
Asuryani: 19
Drukhari: 15
Tyranids: 12
Most points earned per round:
Adeptus Sororitas: 27.34
Dark Angels: 26.50
Imperial Knights: 24.12
Renegade Knights: 24
Ynnari: 23.85
Highest win percentage:
Renegade Knights: 75%
Dark Angels: 75:
Genestealer Cults: 58.33%
Drukhari: 55.32%
Imperial Knights: 55.28%
T’au: 55.07%
found it for you. Taking IG as your primary detachment meant on average you didnt not make the top for Win % or most points per round
Automatically Appended Next Post:
Marmatag wrote: The problem is you can't look at the field and draw a conclusion in regards to balance. And, i've seen that data, it did not account for "percentage used," it just had if a faction was used, at all. You are flatly misrepresenting this data. I would suggest you provide what you have.
You have to filter out the players who lose all, or most of their games. There's a reason people look at the lists that finish in the top 20 or so. Because those lists performed well. I don't care if B. Billy Bumshoes brought a "fluffy" list and got stomped out. That doesn't mean anything in regards to game balance. And the bottom of any tournament has a ton of total crap lists, especially if it's a big event.
You are actively arguing that we should induce survivorship bias into our statistical analysis
Asmodios wrote: A. You would need to make the strategems of equal power for both army A and B for this to become valid.
Yes you would and should! The idea that some armies have better stratagems by design rather than accident has absolutely no evidence. Sure, some armies have better stratagems, but some armies have better units too. It is not because GW intentionally tried to make it so, they just failed at balancing.
Also not sure how you can not agree that soup is inherently more powerful....... look at the number of soup armies in the top of tournaments and the number of mono guard...... there's a reason why all the top players for all the top factions are bringing soup and not mono guard lists. Take one look at nova and explain to me how soup isn't the issue 10/10 of the top lists are soup
"Yes, there is some extreme soup builds that are better than most (or even all) monobuilds, but I'd argue that most soup combinations are actually weaker than mono guard. If you punish soup in general while leaving guard as it is, then it just widens that gap. "
What part of this you didn't understand? Tournaments are representative of the absolutely cutting edge builds, and sure those seem to be some specific soup builds (mainly due poor balancing between codices.) It doesn't change the fact that excluding those couple of anomalies, mono guard is stronger than most soup builds. Do you think Black Templars + Sisters is stronger than mono-Guard? Ad Mech + Inquisition stronger than mono-Guard? No, they aren't. Sure Salamanders + Guard is stronger than pure Salamanders, but that's because Guard is disproportionately strong, and again pure Guard list would still be stronger.
What you are arguing is that mono guard is stronger then soup because of soup builds that we dont see and have no data on. Using that logic i could say that mono guard is awful because everyone we aren't looking at is spamming chimeras and sentinals. We have to work within the data data provided or we are simple just stating opinion with nothing to back up our claims
Nothing interesting about that top 10. It's exactly expected. 3 AM/ 2 Knights. Every single list including knights and AM some with BA - some with Custodes. A few Ynnari. A few Choas soupers. This is exactly expected. Also - top 10 has 3 AM primary...yet - you are trying to make a case that it hurts your win % to do this? Are you arguing in good faith here seriously?
Notice that primary guard did not make the top spots for win percentage or points earned per round. That primary guard preformed bellow mono armies like Tau. primary guard is stronger then mono guard and still finished bellow must take mono armies. This kind of evidence supports my previous claims that wile mono guard is good it is actually not supperior to armies like tau that strangely dont have 3 threads a day about how they need a nerf.
Tau actually have a lot more threads about how much they suck. LOL. I'm not going to argue that tau isn't strong as a mono force. I'd rate them in the top 5 for mono army. I am considering taking Tau to LVO just to see how well I can do. Though - I would really like to have a chance to win it all (even if it is slim) so I will probably be bringing some sort of knight variation depending on what chapter approved does.
Anyways. Tau are certainly a strong army. I win a lot more than I lose with them - and it's usually not close. I felt so bad playing against my good friends mono Custodes army - the only units I lost all game. A single ghost keel (which I kind of let him kill) and 3 fire warriors and 3 drones.
Meanwhile I destroyed a squad of bikes and his super dread (I forget it's name but it's basically a leviathan) and a unit of guards all at -1 to hit the first turn (and I didn't roll very good ether). Basically GG right there.
Marmatag wrote: Except your analysis doesn't take into account the quality of the list. Some lists are bad. Some lists are good. Your analysis conflates lists by Keyword, when in reality armies like Guard, Tau, etc, have a multitude of different ways they can be played. Stop being obtuse.
And again, do you think BCP magically scans the screenshot of a napkin with coffee stains on it where the list was written? It would have to be a player reporting a misrepresented faction for it to be caught. I know a guy who accidentally misreported his faction and it wasn't caught until way later after the event.
1. They covered how the data was gathered in that weeks FLG podcast. They took every list plugged it into excel and ran that stats
2. Once again you skew the data by removing the left side of the bell curve. For instance only look at the undefeated DG player from BAO... remove all other DG players..... ZOMG DG has 100% win rate they cannot be beaten nerf blightlord terminators now!!!! skewing the data on purpose gives you a poor view of win percentages. There is a reason why no statistical analysis does this. I would have failed every stats class i took in college had i run regressions this way
Most statistical analysis removes outliers. That gives you the best idea of the averages. We don't to know whos average - we want to know whos best. So you should remove the bottom 50% I say only look at the top 20%. The trash armies at the bottom offer nothing to the question of what is best.
That's not "removing an outlier" and my post above shows why it would incredibly skew data in such a small sample size. For example, the DG would shoot up to a 100% win rate. To do it correctly too you would need to repeat the exact tournament with the same opponents over and over while having them play all available army so that way you could get a true "player skill". Even then under perfect conditions, you are changing the definition of "best army/ faction" To "best army/ faction while being played by top 10%" meaning that the data would really only be useful when taken by top 10-20%. You also run into the issue of the top 10-20% not even bringing certain factions so losing data on where those fall.
At the end of the day skewing this sample in any of those ways is going to give you a worse picture of the data.
[/spoiler]
Wouldn't it be better to cut the lowest part of the tournament results AND the top tournament results? I mean, i'm an engineer not a math expert, but that's what i would do.
[/spoiler]
Yes if you wanted to cut off the tail ends of the bell curve on both sides it would give better results then simply cutting off the one end that doesnt fit your narrative.
For example, if you wanted to know the average physical strength of men vs women you might want to exclude something like "pro athletes" as you could argue this is not only a poor example of "average" but men might have a much larger pool of pro athletes. On the bottom end you might want to exclude something like physically disabled for the same reason.
This message was edited 1 time. Last update was at 2018/10/04 21:20:14
I'm not interested in discussing poorly chosen analogies.
We all understand what's happening here, and what you're doing: You deflate the success rate of good armies by including bad lists. That's your entire post chain on this topic here, in a nutshell.
This message was edited 1 time. Last update was at 2018/10/04 21:23:40
Galas wrote: I remember when Marmatag was a nooby, all shiney and full of joy. How playing the unbalanced mess of Warhammer40k in a ultra-competitive meta has changed you
Bharring wrote: He'll actually *change his mind* in the presence of sufficient/sufficiently defended information. Heretic.
Marmatag wrote: Because we're looking at the potential of each army with the best configuration(s).
No one cares you can run sub-optimal lists. I'll openly acknowledge every codex can suck if you screw it up hard enough. But that's not the point, because the game isn't about racing to the bottom.
Not everyone is concerned about the top 10% though.
Some of us are more concerned about what happens if we bring our favourite army to a pick up game at our local store.
Then you should find that out by doing it, not by hunting for broken combos on the internet. Netlisting sucks especially in a small pond meta.
What has that got to do with looking at the performance data for mid level play?
