Author |
Message |
 |
|
 |
Advert
|
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. |
|
 |
![[Post New]](/s/i/i.gif) 2023/01/23 23:33:47
Subject: Re:Prediction Time
|
 |
Perfect Shot Ultramarine Predator Pilot
|
alextroy wrote:There are only two ways two players can go 10-0 and 0-10 with the same Faction playing the same Tournament Rules:
One Player has a crap list while the other has a fantastic list
One player is a crap player and the other is a fantastic player
Yes, that's exactly the point. Overall faction win rates tell you very little about the health of the metagame because they can't tell the difference between a faction where every list is at about 50% win rate and a faction where one list is dominating but every other list sucks.
The reason GW focuses on faction win rates instead of more advanced metrics is that it's the easiest data to obtain and by ignoring internal balance issues entirely they can post smug self-congratulatory articles about how great everything is even if 90% of a codex is F-tier garbage.
So, yes. It is fine to call that a 50% win rate.
No it isn't. That's a single blatantly overpowered list being brought down to "fair" levels by including a weaker list/player.
|
This message was edited 1 time. Last update was at 2023/01/23 23:35:02
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 00:23:05
Subject: Re:Prediction Time
|
 |
Confessor Of Sins
|
I guess you don't listen to their Metawatch videos. They are also looking at internal balance along with win rate. They want to increase the variety of units being included in list. The AOS videos are more explicit about it, but there is enough talk in the 40K videos to make me think they are looking at unit inclusion rates to determine how to adjust unit points values in those updates.
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 00:37:16
Subject: Re:Prediction Time
|
 |
Owns Whole Set of Skullz Techpriests
Versteckt in den Schatten deines Geistes.
|
Only if I'm feeling down and need a good laugh.
|
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 02:12:00
Subject: Re:Prediction Time
|
 |
Longtime Dakkanaut
|
alextroy wrote:I guess you don't listen to their Metawatch videos. They are also looking at internal balance along with win rate. They want to increase the variety of units being included in list. The AOS videos are more explicit about it, but there is enough talk in the 40K videos to make me think they are looking at unit inclusion rates to determine how to adjust unit points values in those updates.
You mean the unit points like getting rid of wargear costs?
Yes, that takes a lot of skill.
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 04:04:51
Subject: Re:Prediction Time
|
 |
Stubborn Dark Angels Veteran Sergeant
Vancouver, BC
|
Aecus Decimus wrote:No it isn't. That's a single blatantly overpowered list being brought down to "fair" levels by including a weaker list/player.
How do you propose we remove that noise from that data?
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 04:30:59
Subject: Re:Prediction Time
|
 |
Perfect Shot Ultramarine Predator Pilot
|
Canadian 5th wrote:Aecus Decimus wrote:No it isn't. That's a single blatantly overpowered list being brought down to "fair" levels by including a weaker list/player.
How do you propose we remove that noise from that data?
https://www.goonhammer.com/hammer-of-math-tournaments-in-winning-position-and-other-key-performance-stats-explained/
TIWP, win path, etc, allow you to sort out these things. A polarized faction between 10-0 and 0-10 will have a 50% win rate but a high TIWP, reflecting the fact that if you avoid being a loser you're going to consistently be in position to win the tournament. Similarly, a faction with a low TIWP but a 70% win rate is cleaning up the losers bracket but gets promptly booted out of contention for the top tables once it hits a good opponent. And because major events usually require online list submission you can look at the data at a finer resolution than entire factions. You can look at win rates/TIWP/etc for each sub-faction or list archetype, and you can look at usage rates for each unit in the codex to understand the state of internal balance. That 10-0/0-10 faction would be easy to handle if you look at the data and see that one archetype is going 10-0 and the other archetypes are 0-10, meaning you have a few overpowered outlier units in an otherwise very weak faction.
|
This message was edited 1 time. Last update was at 2023/01/24 04:31:40
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 06:00:09
Subject: Re:Prediction Time
|
 |
Stubborn Dark Angels Veteran Sergeant
Vancouver, BC
|
Aecus Decimus wrote: Canadian 5th wrote:Aecus Decimus wrote:No it isn't. That's a single blatantly overpowered list being brought down to "fair" levels by including a weaker list/player.
How do you propose we remove that noise from that data?
https://www.goonhammer.com/hammer-of-math-tournaments-in-winning-position-and-other-key-performance-stats-explained/
TIWP, win path, etc, allow you to sort out these things. A polarized faction between 10-0 and 0-10 will have a 50% win rate but a high TIWP, reflecting the fact that if you avoid being a loser you're going to consistently be in position to win the tournament. Similarly, a faction with a low TIWP but a 70% win rate is cleaning up the losers bracket but gets promptly booted out of contention for the top tables once it hits a good opponent. And because major events usually require online list submission you can look at the data at a finer resolution than entire factions. You can look at win rates/TIWP/etc for each sub-faction or list archetype, and you can look at usage rates for each unit in the codex to understand the state of internal balance. That 10-0/0-10 faction would be easy to handle if you look at the data and see that one archetype is going 10-0 and the other archetypes are 0-10, meaning you have a few overpowered outlier units in an otherwise very weak faction.
That still abstracts a lot, has no way of accounting for the differences in player skill between different tournaments that don't share the same player pool, and generates a minuscule data pool compared to both the complexity of the system and the total player base. There's a reason I went so hard on suggesting a rules complete, GW developed, VTT with things like skill-based matchmaking as a means of generating a useful level of balancing data. As it is, GW can and should do better but I doubt tournament data is doing anything that a reasonable level of prerelease couldn't accomplish.
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 11:32:27
Subject: Prediction Time
|
 |
Longtime Dakkanaut
|
If there is a single overpowered list you'd expect to see it everywhere very quickly. This is why placings are generally more of a measure than faction win%. But they tend to go together.
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 14:45:24
Subject: Prediction Time
|
 |
Stubborn Dark Angels Veteran Sergeant
Vancouver, BC
|
Tyel wrote:If there is a single overpowered list you'd expect to see it everywhere very quickly. This is why placings are generally more of a measure than faction win%. But they tend to go together.
That only tells you that a type of list is broken. It doesn't tell you why it's broken. Like AdMech was mostly broken because it had a high damage first round that was difficult to hide from, did the data capture that?
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 16:03:15
Subject: Prediction Time
|
 |
Longtime Dakkanaut
|
Lists tend to get published and then copied. So... yes? Not really sure what you mean. Not sure how many people were going "flyers, Ballistarii and big blocks of boosted up Skitarii are fine. Nerf the Kataphron instead."
By and large - and certainly compared to the situation pre 8th - GW at least targets the right things for buffs/nerfs.
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 16:31:31
Subject: Prediction Time
|
 |
Longtime Dakkanaut
|
Canadian 5th wrote:Tyel wrote:If there is a single overpowered list you'd expect to see it everywhere very quickly. This is why placings are generally more of a measure than faction win%. But they tend to go together.
That only tells you that a type of list is broken. It doesn't tell you why it's broken. Like AdMech was mostly broken because it had a high damage first round that was difficult to hide from, did the data capture that?
Depends what you mean by "data". Strictly the various stats that track overall army performance? Probably not, but most tournaments require online list submission and those lists are available for people to view. The reality is, if you want to get a good idea of the balance issues of any given faction you need to do a deeper dive than just looking at some high-level stats. They can help guide your approach but there's no substitute for doing a proper in-depth analysis of the lists that are winning tournaments. I'd expect GW to be able to do much more in this area than they currently are.
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 17:01:29
Subject: Prediction Time
|
 |
Stubborn Dark Angels Veteran Sergeant
Vancouver, BC
|
Tyel wrote:Lists tend to get published and then copied. So... yes? Not really sure what you mean. Not sure how many people were going "flyers, Ballistarii and big blocks of boosted up Skitarii are fine. Nerf the Kataphron instead."
By and large - and certainly compared to the situation pre 8th - GW at least targets the right things for buffs/nerfs.
It means that you don't have any data on the second-order effects your balance changes might have. Look at what happened to DE with each round of nerfs just revealing the new best unit until they were finally dethroned by the new codices arriving. Nor does it give you any data on what else might rise up in the meta once the current top list is no longer holding down other lists that may well be potent in their own right.
Slipspace wrote: Canadian 5th wrote:Tyel wrote:If there is a single overpowered list you'd expect to see it everywhere very quickly. This is why placings are generally more of a measure than faction win%. But they tend to go together.
That only tells you that a type of list is broken. It doesn't tell you why it's broken. Like AdMech was mostly broken because it had a high damage first round that was difficult to hide from, did the data capture that?
Depends what you mean by "data". Strictly the various stats that track overall army performance? Probably not, but most tournaments require online list submission and those lists are available for people to view. The reality is, if you want to get a good idea of the balance issues of any given faction you need to do a deeper dive than just looking at some high-level stats. They can help guide your approach but there's no substitute for doing a proper in-depth analysis of the lists that are winning tournaments. I'd expect GW to be able to do much more in this area than they currently are.
Yeah, but even that deep dive is only educated guessing without a pool of data to draw from. The disadvantage that GW has is that they simply have no way to gather good data and don't have any way to simulate games in such a way as to gather that data. Outside of using some method to poll many more games than have ever been played over the history of 40k tournaments, the error bars on that data they do get are so large as to make all but the highest-level observations meaningless. I think that's why a VTT with online multiplayer game level of matchmaking and data collection could vastly improve the balance of the game. Beyond that, a purpose-built AI running on high-end hardware could generate the required data with a known skill level but depending on how it was trained that data may have issues of its own that could take a while to parse.
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 18:02:36
Subject: Prediction Time
|
 |
Decrepit Dakkanaut
|
Firstly, this is a garbage narrative.
**Unless you're a playtester, in which case you are encouraged to lobby GW to make your personal army win more as long as you make up for it by nerfing some of the units you don't use
Aecus Decimus wrote: H.B.M.C. wrote:But as long as their win rate sits in that magical 45%-55% Goldilocks zone, everything is A-Ok, right?
Exactly! So what if you go 0-10, some random player on the other side of the world went 10-0 spamming TH/ SS terminators and the two of you combined were a perfect 50%. It's the faction as a whole* that matters, not your individual experiences**.
You guys are attempting to make a bit of a strawman here. Making sure factions are in the 45 to 55% band is step 1. Reworking internal balance is a gradual step 2. It takes a lot of analysis to take lists, review them, determine their strength, and ponder why they won or lost against a particular list. And access to that data in a format that is processable is non-existent at the moment.
What list beat that DA list?
Aleya
Trajann
SC on Bike
3 Allarus
2x1 Allarus
2x10 Wardens
Vexilus
5 Witchseekers
Rhino
More difficult to remove models.
Here's another DA player that failed even with lots of terminators:
Talonmaster
5 Heavy Intercessors
2x5 Infiltrators
BG Ancient
RW Apoth
1x2 Command
2x10 Termies
2 AB w/ MM
2x3 Outriders
Thunderstrike
1x3 Eradicators
Are they a bad player or did the internal selections weaken the list enough? Their record is 3-5 so surely they must have been terrible, right?
L - 79 to 85, CK
L - 72 to 83, CK
L - 80 to 81, Necrons
W - 85 to 61, Necrons
L - 57 to 91, Nids
L - 39 to 82, Night Lords
W - 93 to 57, Necrons
W - 92 to 53, Votann
This player has a 3-5 record, but 3 of those loses were quite close to wins. Was it the opponents or the lists who were better? Or both? Was it a quirk of the mission or terrain?
Losing the first 3 rounds is going to put you into lower brackets and potentially easier opponents, but most of the lists they faced were pretty reasonable. Was the player pool of a higher caliber at this tournament?
The 7-1 DA player had the following results:
W - 95 to 28, Custodes + Wardogs
W - 76 to 64, Craftworld
W - 97 to 77, Orks
W - 85 to 75, Blood Angels
W - 91 to 90, CK
W - 75 to 71, Space Wolves ( 3x Fangs, 9 Inceptors )
W - 82 to 80, Tau ( 2 HH, 2 Riptides )
L - 65 to 85, Custodes
The vast majority of these games are not blow-outs, either. They barely beat Knights here.
The win rates are riding a razor's edge. Your guy who goes 0-10 and loses by 5 to 10 points every game is not the same as going 0-10 and losing by 30+ a game.
|
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 18:31:03
Subject: Prediction Time
|
 |
Stubborn Dark Angels Veteran Sergeant
Vancouver, BC
|
This.
Though I'd even argue that a 45-55% win rate bracket is far too large. People will bring up Chess, while ignoring that it is best played in larger sets with players each taking both black and white an equal number of times. Go has an entire system of handicapping to balance player strength and first-turn advantage. There's also a need factor in perceived strength, how enjoyable a list is to play and play against, faction and list diversity, how different lists perform at different levels of player skill, how board layouts change win-rates, how objectives change things, etc.
In short the data GW can collect cannot meaningfully address game balance for a game like 40k.
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 18:55:10
Subject: Prediction Time
|
 |
Bounding Dark Angels Assault Marine
|
I believe balance could be improved if the rules writers showed some restrained concerning special rules that ignore special rules.
Additionally, has anyone made a graph plotting unit average damage and speed vs cost? There are many internal balance issues that could be fixed looking at such a graph.
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 19:13:58
Subject: Prediction Time
|
 |
Longtime Dakkanaut
|
Brickfix wrote:I believe balance could be improved if the rules writers showed some restrained concerning special rules that ignore special rules.
Additionally, has anyone made a graph plotting unit average damage and speed vs cost? There are many internal balance issues that could be fixed looking at such a graph.
We have very few cases of rules ignoring other rules, and all the existing ones have been there since the first dexes.
People like to whine about power creep, but the very first dex already had a unit which ignored all invul AND all wound gates.
The only "ignore rule" which was added later, was the FAQ to allow the GK warlord trait to ignore the demon saves... because demon saves were not there yet.
So no, that is definitely a factor nor an issue.
Plotting stats gives you very limited info.
The problematic units are almost always a result of stacking of rules, very rarely they are stat issues. When they are, the fix is usually very fast, since just increasing the points solves the issue.
Units not taken in a codex are usually not taken due to the following factors:
1) It is an HQ and another HQ offers better synergy with whatever is meta.
2) The unit is too hard to use efficiently/ a competent opponent can neutralize them too easily.
3) The unit has a too narrow application/another unit covers the same role but with a broader applicability
4) The unit isn't good in the current meta
Apart from some limited case in Space Marines, you almost never have a case where unit x is unit y but better. It is always a matter of how you can use them on the table compared to the alternatives.
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 20:22:56
Subject: Prediction Time
|
 |
Loyal Necron Lychguard
|
Spoletta wrote:Units not taken in a codex are usually not taken due to the following factors:
1) It is an HQ and another HQ offers better synergy with whatever is meta.
2) The unit is too hard to use efficiently/ a competent opponent can neutralize them too easily.
3) The unit has a too narrow application/another unit covers the same role but with a broader applicability
4) The unit isn't good in the current meta
That's just rationalising after the fact. The only thing that actually matters is pts cost. Otherwise we'd never see a Terminator be better in tournaments than a Knight. Because 10 Knights beat 10 Terminators. HQs can be efficient despite a lack of synergy. Despite having counterplay a unit can still be overpowered. Enough melta shots will kill a squad of Ork Boyz. The only thing that matters is points. You can analyze why some units are under or overcosted but if you can't say that every HQ unit which is an inferior option to another HQ is less competitive then your analysis doesn't really tell us anything. You can say that 5 point Plaguebearers have a good damage output, are monstrously durable which lets them hold objectives and they can screen out the table, while 15 point Plaguebearers hit like wet noodles, die to a stiff breeze and don't take up much space, but they're the same unit, the only difference is points. You can calculate how good combos are as well, like comparing a buff castle to units that go on their own and figure out how much effect the +1D Strat has when it's combined with re-rolls to hit and to wound as well as extra AP and get an expected number of extra wounds caused to various targets. GW should absolutely do the basic math on their game.
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 20:32:41
Subject: Prediction Time
|
 |
Humming Great Unclean One of Nurgle
|
vict0988 wrote:Spoletta wrote:Units not taken in a codex are usually not taken due to the following factors:
1) It is an HQ and another HQ offers better synergy with whatever is meta.
2) The unit is too hard to use efficiently/ a competent opponent can neutralize them too easily.
3) The unit has a too narrow application/another unit covers the same role but with a broader applicability
4) The unit isn't good in the current meta
That's just rationalising after the fact. The only thing that actually matters is pts cost. Otherwise we'd never see a Terminator be better in tournaments than a Knight. Because 10 Knights beat 10 Terminators. HQs can be efficient despite a lack of synergy. Despite having counterplay a unit can still be overpowered. Enough melta shots will kill a squad of Ork Boyz. The only thing that matters is points. You can analyze why some units are under or overcosted but if you can't say that every HQ unit which is an inferior option to another HQ is less competitive then your analysis doesn't really tell us anything. You can say that 5 point Plaguebearers have a good damage output, are monstrously durable which lets them hold objectives and they can screen out the table, while 15 point Plaguebearers hit like wet noodles, die to a stiff breeze and don't take up much space, but they're the same unit, the only difference is points. You can calculate how good combos are as well, like comparing a buff castle to units that go on their own and figure out how much effect the +1D Strat has when it's combined with re-rolls to hit and to wound as well as extra AP and get an expected number of extra wounds caused to various targets. GW should absolutely do the basic math on their game.
Points matter, 100%.
But rules and statlines do too.
Here's a challenge-I want to assign a points cost to the following unit:
I will freely admit that that specific unit is stupid and would never see print. But the concept, that of a glass cannon unit, is so common that it has its own term.
|
Clocks for the clockmaker! Cogs for the cog throne! |
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 20:52:56
Subject: Prediction Time
|
 |
Decrepit Dakkanaut
|
Brickfix wrote:I believe balance could be improved if the rules writers showed some restrained concerning special rules that ignore special rules.
Additionally, has anyone made a graph plotting unit average damage and speed vs cost? There are many internal balance issues that could be fixed looking at such a graph.
The marine bench is stupid deep so good luck making everything in that book coherent.
In any case someone might take 3 MM Attack Bikes for 150, because they can get within 12" without allowing an enemy response, but they have fewer shots than a full MM Dev squad with cherub. That Dev Squad is a lot softer and can't get in half range very easily so you're probably forking over another 70 for a pod. You risk getting hit coming out of the pod and the opponent could put chaff in the way to keep you out of half range, but at least you can come down turn 1.
Additionally, you could take a second dev squad of Grav. Now you can pick which squad goes in the pod depending on the opponent, but you have more soft models and a unit that has to walk.
Devs can action and auspex. Bikes are a better target for combat revival and can pick up a -1 to be hit.
There is no calculation that you can apply to value devastators against attack bikes and come to a reasonable conclusion.
|
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 20:56:21
Subject: Prediction Time
|
 |
Huge Hierodule
Mexico
|
JNAProductions wrote:
Enemy units have +1 to-hit this unit.[/spoiler]I will freely admit that that specific unit is stupid and would never see print. But the concept, that of a glass cannon unit, is so common that it has its own term.
While I agree that rules and stats matter, are you seriously suggesting that you cannot give appropriate point costs to glass cannon units?
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 21:08:11
Subject: Prediction Time
|
 |
Humming Great Unclean One of Nurgle
|
If they’re sufficiently glassy and cannony… yes.
If you go first with a glass cannon list, you win.
If you go second, you lose.
I don’t think any current armies are quite that extreme in the fragility and offensive power, though.
|
Clocks for the clockmaker! Cogs for the cog throne! |
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 21:32:42
Subject: Prediction Time
|
 |
Discriminating Deathmark Assassin
|
Daedalus81 wrote:Firstly, this is a garbage narrative.
**Unless you're a playtester, in which case you are encouraged to lobby GW to make your personal army win more as long as you make up for it by nerfing some of the units you don't use
Aecus Decimus wrote: H.B.M.C. wrote:But as long as their win rate sits in that magical 45%-55% Goldilocks zone, everything is A-Ok, right?
Exactly! So what if you go 0-10, some random player on the other side of the world went 10-0 spamming TH/ SS terminators and the two of you combined were a perfect 50%. It's the faction as a whole* that matters, not your individual experiences**.
You guys are attempting to make a bit of a strawman here. Making sure factions are in the 45 to 55% band is step 1. Reworking internal balance is a gradual step 2. It takes a lot of analysis to take lists, review them, determine their strength, and ponder why they won or lost against a particular list. And access to that data in a format that is processable is non-existent at the moment.
What list beat that DA list?
Aleya
Trajann
SC on Bike
3 Allarus
2x1 Allarus
2x10 Wardens
Vexilus
5 Witchseekers
Rhino
More difficult to remove models.
Here's another DA player that failed even with lots of terminators:
Talonmaster
5 Heavy Intercessors
2x5 Infiltrators
BG Ancient
RW Apoth
1x2 Command
2x10 Termies
2 AB w/ MM
2x3 Outriders
Thunderstrike
1x3 Eradicators
Are they a bad player or did the internal selections weaken the list enough? Their record is 3-5 so surely they must have been terrible, right?
L - 79 to 85, CK
L - 72 to 83, CK
L - 80 to 81, Necrons
W - 85 to 61, Necrons
L - 57 to 91, Nids
L - 39 to 82, Night Lords
W - 93 to 57, Necrons
W - 92 to 53, Votann
This player has a 3-5 record, but 3 of those loses were quite close to wins. Was it the opponents or the lists who were better? Or both? Was it a quirk of the mission or terrain?
Losing the first 3 rounds is going to put you into lower brackets and potentially easier opponents, but most of the lists they faced were pretty reasonable. Was the player pool of a higher caliber at this tournament?
The 7-1 DA player had the following results:
W - 95 to 28, Custodes + Wardogs
W - 76 to 64, Craftworld
W - 97 to 77, Orks
W - 85 to 75, Blood Angels
W - 91 to 90, CK
W - 75 to 71, Space Wolves ( 3x Fangs, 9 Inceptors )
W - 82 to 80, Tau ( 2 HH, 2 Riptides )
L - 65 to 85, Custodes
The vast majority of these games are not blow-outs, either. They barely beat Knights here.
The win rates are riding a razor's edge. Your guy who goes 0-10 and loses by 5 to 10 points every game is not the same as going 0-10 and losing by 30+ a game.
Hold on, I'm super confused about scoring now.
I was told that if there's less than a 5 point difference, the game is counted as a draw.
|
213PL 60PL 12PL 9-17PL
(she/her) |
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 21:37:59
Subject: Prediction Time
|
 |
Longtime Dakkanaut
|
Nope, the draw happens only if the scores are identical.
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 21:42:12
Subject: Prediction Time
|
 |
Decrepit Dakkanaut
|
Blndmage wrote:
Hold on, I'm super confused about scoring now.
I was told that if there's less than a 5 point difference, the game is counted as a draw.
I imagine there are some TOs out there who might do that, but I have never seen it in the wild. I am not well informed on if that would be a good idea for wider use.
|
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 23:00:10
Subject: Re:Prediction Time
|
 |
Crazed Spirit of the Defiler
|
The 20-0 scoring format used by WTC and others and recently adopted in some GW tournaments requires a win by a margin of 6 VP to get an 11-9 victory instead of a 10-10 draw.
Just thought I would mention that as one of the few people in the thread titled "Prediction Time" who was willing to try to guess which the best faction would be, I did ok https://www.dakkadakka.com/dakkaforum/posts/list/808330.page#11474159
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/24 23:25:56
Subject: Re:Prediction Time
|
 |
Discriminating Deathmark Assassin
|
EightFoldPath wrote:The 20-0 scoring format used by WTC and others and recently adopted in some GW tournaments requires a win by a margin of 6 VP to get an 11-9 victory instead of a 10-10 draw.
Can you expand on that?
I'm finding the competitive scene far less monolithic in terms of rules. It's really confusing.
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/25 05:11:15
Subject: Prediction Time
|
 |
Loyal Necron Lychguard
|
JNAProductions wrote: vict0988 wrote:Spoletta wrote:Units not taken in a codex are usually not taken due to the following factors:
1) It is an HQ and another HQ offers better synergy with whatever is meta.
2) The unit is too hard to use efficiently/ a competent opponent can neutralize them too easily.
3) The unit has a too narrow application/another unit covers the same role but with a broader applicability
4) The unit isn't good in the current meta
That's just rationalising after the fact. The only thing that actually matters is pts cost. Otherwise we'd never see a Terminator be better in tournaments than a Knight. Because 10 Knights beat 10 Terminators. HQs can be efficient despite a lack of synergy. Despite having counterplay a unit can still be overpowered. Enough melta shots will kill a squad of Ork Boyz. The only thing that matters is points. You can analyze why some units are under or overcosted but if you can't say that every HQ unit which is an inferior option to another HQ is less competitive then your analysis doesn't really tell us anything. You can say that 5 point Plaguebearers have a good damage output, are monstrously durable which lets them hold objectives and they can screen out the table, while 15 point Plaguebearers hit like wet noodles, die to a stiff breeze and don't take up much space, but they're the same unit, the only difference is points. You can calculate how good combos are as well, like comparing a buff castle to units that go on their own and figure out how much effect the +1D Strat has when it's combined with re-rolls to hit and to wound as well as extra AP and get an expected number of extra wounds caused to various targets. GW should absolutely do the basic math on their game.
Points matter, 100%.
But rules and statlines do too.
Here's a challenge-I want to assign a points cost to the following unit:
I will freely admit that that specific unit is stupid and would never see print. But the concept, that of a glass cannon unit, is so common that it has its own term.
*Create model for Super Annihilator Weapons Platform.
*Write fluff that is fitting for the model.
*Draft as many versions of alpha rules as possible.
*Get initial impressions from any playtester that is interested.
*Create 3 different beta versions based on feedback.
*The casual playtest group find out which version is more fun and fluffy on the table.
*Finalize rules and calculate how hard it is to kill with various weapons and how much of various units it kills, this gives the unit a point range where it is potentially balanced.
*The competitive playtest group spam the unit in at least 3 games and the mathematically derived points value is adjusted within the calculated range.
*The book is printed after all of this is done.
I'd estimate 4500 points based on how much it can kill each turn regardless of LOS and it's fragility. But I explained above how I'd want GW to write and playtest rules.
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/25 05:20:00
Subject: Re:Prediction Time
|
 |
Perfect Shot Ultramarine Predator Pilot
|
Blndmage wrote:EightFoldPath wrote:The 20-0 scoring format used by WTC and others and recently adopted in some GW tournaments requires a win by a margin of 6 VP to get an 11-9 victory instead of a 10-10 draw.
Can you expand on that?
I'm finding the competitive scene far less monolithic in terms of rules. It's really confusing.
The point is that a narrow win is treated as a draw for scoring purposes, the assumption being that if you only win by 1 VP it was pretty much random chance who was going to win and it's scored as a draw.
The scoring system used for recording match results uses a 20 point scale instead of straight win/loss/draw. The winner is awarded between 11 and 20 points based on how large the margin of victory was, the loser gets 20 minus the winner's score. If the margin of victory is 6 VP or less each player gets 10 points. Your total on the 0-20 scale is used for pairings each round and determining the event winner. Automatically Appended Next Post: vict0988 wrote:I will freely admit that that specific unit is stupid and would never see print.
There you go, now you admit that points are not 100% of balance and "the only thing that actually matters is pts cost" was a false statement. Automatically Appended Next Post:
It is not. Remember, 40k has to cope with a bad list factor that doesn't apply in other games. In chess you don't have players taking a list with nothing but pawns because "my lore is that we don't have a queen", in 40k you absolutely have people showing up with bad lists based on their lore. Even if you have a perfectly balanced game at the top tables you're going to have a lot of variability in win rates simply because of random luck involving who showed up on the lower tables. If you try to set a narrower tolerance you significantly increase the risk of over-correcting and "fixing" a problem that never existed, making balance worse than if you'd just accepted a 46% win rate.
And yes, I know you like the idea of a virtual tabletop version of 40k. It isn't going to happen for a variety of reasons and there's no point in discussing how it could be used for balance data.
|
This message was edited 2 times. Last update was at 2023/01/25 05:27:46
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/25 05:45:35
Subject: Re:Prediction Time
|
 |
Stubborn Dark Angels Veteran Sergeant
Vancouver, BC
|
Aecus Decimus wrote:It is not. Remember, 40k has to cope with a bad list factor that doesn't apply in other games. In chess you don't have players taking a list with nothing but pawns because "my lore is that we don't have a queen"
If the game was balanced replacing certain meta units with more troops shouldn't overly disadvantage you as troops should have a valuable role that other units can't fill. So really, that shouldn't matter even if we expect such lists to be in our current dataset often enough to be more than noise.
|
|
 |
 |
![[Post New]](/s/i/i.gif) 2023/01/25 06:02:41
Subject: Re:Prediction Time
|
 |
Loyal Necron Lychguard
|
Aecus Decimus wrote: vict0988 wrote:I will freely admit that that specific unit is stupid and would never see print. There you go, now you admit that points are not 100% of balance and "the only thing that actually matters is pts cost" was a false statement. JNAProductions wrote:I will freely admit that that specific unit is stupid and would never see print. But the concept, that of a glass cannon unit, is so common that it has its own term.
You got your quotes mixed up  The only thing that for determining whether a unit is OP or UP is pts cost. A unit can be badly designed if it isn't fun to play with or against or if it doesn't represent the fluff of the unit. The two aren't linked. A well-designed unit can be OP if it costs too little and UP if it costs too much, a badly designed unit like JNA's Super Annihilator Weapons Platform can be OP if it costs 100 points and UP if it costs 100000 points.
|
This message was edited 1 time. Last update was at 2023/01/25 06:03:26
|
|
 |
 |
|