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.

 Galas wrote:
W/L should be considered a secondary factor unless it is too low or too high. Once we reach 40% or get past 57-58% , thats an obvious problem (Unless we have a data of 5-6 lists played)


I would not say top 4-8 is very good.

For example, is the top 10 of a 60 man tournament as relevant as the top 50 of a 700 person tournament like adepticon?

In general, the first third of any tournament is were the "competitive" people will end up. Thats the relevant data we should measure.

This is the same issue I have with NCAA Football rankings.
Teams that go unbeaten often wind up ranked lower than teams with one or two losses because Team X "had a tougher schedule". Which is usually due to teams they frequently play getting ranked higher simply because they play Team X.

Talking of Faction Levels,
Win/Loss should have more impact when determining power level. Finishing in the top 10 of a tourney is going to have an effect due to most are going to be entering the third round with 0-1 losses anyway. If they play 10 games a weekend by the time the final matches are announced many will have a 9-0 or 8-1 record. Yeah, the most competitive players will likely be there, but their faction could have anywhere between a 70-30% and a 30-70% win/loss record. On the lower end, the Top 10 player would be the outlier, likely having an optimized list, some serious good luck, more skill than his opponents, and/or favorable matchups prior to the final day (example: an optimized Ork list facing multiple unoptimized Guard lists.)

It is a little of both I'd say.

Both are useful statistics and it depends on what you want out of them.
If you're looking for peak performance IMO you need to be looking at the top ranks so you avoid the noise of casual participants.

My (considerable) gut feeling is that, at least for "competitive" play, higher finishs outweighs w/l as you need good wins rather than just wins, e.g. If a lists constantly wins but only scores 50 each game it'll lose out in the cut to lists going 70+, of course down the FLGS or club a win is a win regardless

Context is critical when analyzing any data like this. Using either of these without context won't help in any meaningful way.
But one idea that I've been liking more and more as a general indicator is combining top finishes with list diversity IE. How many different lists from a faction placed in the top (or high) slots?

A faction that has one list that is hyper-efficient will eventually see that list teched against, and it has no real variety to fall back on.
A faction that has multiple effective lists placing top generally means a healthy/competitive faction. Even if one more popular list being teched against, there is still enough variety that it can be worked around.

 Turnip Jedi wrote:
My (considerable) gut feeling is that, at least for "competitive" play, higher finishs outweighs w/l as you need good wins rather than just wins, e.g. If a lists constantly wins but only scores 50 each game it'll lose out in the cut to lists going 70+, of course down the FLGS or club a win is a win regardless

So you are saying that a "quality win" is tabling a weak opponent as opposed to winning a full turn game by a few points?
That's the problem with a lot of tournaments. Players with "Big Name" recognition never seem to face each other early. So, they rack up quick, easy wins over weaker, less experienced or prepared opponents, garnering as many points as possible so the final day can have "Big Matches".
Those are not "quality wins". Those are like a pro sports team playing a high school team.

Or this... [youtube] https://www.youtube.com/watch?v=vKQtN0YQtC4 [/youtube]

One wonders how things would work out if the "Big Names" played each other on day one.

 helgrenze wrote:

 Turnip Jedi wrote:
My (considerable) gut feeling is that, at least for "competitive" play, higher finishs outweighs w/l as you need good wins rather than just wins, e.g. If a lists constantly wins but only scores 50 each game it'll lose out in the cut to lists going 70+, of course down the FLGS or club a win is a win regardless

So you are saying that a "quality win" is tabling a weak opponent as opposed to winning a full turn game by a few points?
That's the problem with a lot of tournaments. Players with "Big Name" recognition never seem to face each other early. So, they rack up quick, easy wins over weaker, less experienced or prepared opponents, garnering as many points as possible so the final day can have "Big Matches".
Those are not "quality wins". Those are like a pro sports team playing a high school team.

Or this... [youtube] https://www.youtube.com/watch?v=vKQtN0YQtC4 [/youtube]

One wonders how things would work out if the "Big Names" played each other on day one.

Most sports leagues place top seeds versus bottom seeds in the early rounds. Everything from Basketball to Tennis works this way, why shouldn't 40k?

 helgrenze wrote:

 Turnip Jedi wrote:
My (considerable) gut feeling is that, at least for "competitive" play, higher finishs outweighs w/l as you need good wins rather than just wins, e.g. If a lists constantly wins but only scores 50 each game it'll lose out in the cut to lists going 70+, of course down the FLGS or club a win is a win regardless

So you are saying that a "quality win" is tabling a weak opponent as opposed to winning a full turn game by a few points?
That's the problem with a lot of tournaments. Players with "Big Name" recognition never seem to face each other early. So, they rack up quick, easy wins over weaker, less experienced or prepared opponents, garnering as many points as possible so the final day can have "Big Matches".
Those are not "quality wins". Those are like a pro sports team playing a high school team.

Or this... [youtube] https://www.youtube.com/watch?v=vKQtN0YQtC4 [/youtube]

One wonders how things would work out if the "Big Names" played each other on day one.

Well since it's rigged we'll never know.

ccs wrote:
Well since it's rigged we'll never know.

How is seeding top-rated players rigging anything? Every sports league does this, and some of them even reseed in later rounds to preserve the advantage for the higher-seeded teams. Nobody claims that this is rigging the playoffs/tournament when sports leagues do it so why is it rigging when 40k does it?

 Canadian 5th wrote:

Most sports leagues place top seeds versus bottom seeds in the early rounds. Everything from Basketball to Tennis works this way, why shouldn't 40k?

That's rigging the odds. Sports teams dont PURPOSELY pit top seed against bottom seed - it just happens that way because most sports are divided into leagues (there by creating sort of hierarchy that may not be reflective of the teams' ranking overall. Bottom ranking team from one league may be better than 1st seed of another league) that come together at the end in tournament format. Not to mention the tournament rosters are more or less predetermined, and teams work during the regular season to secure a particular spot in the bracket, so unless you can implement a system where players work on their ranking prior tp the tournament, youre rigging the game so that top players will always have better chance for top place.

 Canadian 5th wrote:

ccs wrote:
Well since it's rigged we'll never know.

How is seeding top-rated players rigging anything? Every sports league does this, and some of them even reseed in later rounds to preserve the advantage for the higher-seeded teams. Nobody claims that this is rigging the playoffs/tournament when sports leagues do it so why is it rigging when 40k does it?

That is usually after a season when the win/loss record is built. You don't see the NFL deciding that the winner of the Super Bowl won't play any team that had a winning record.
In fact, when making the schedule, teams play their division rivals (twice), one division from their conference, one from the other, and the teams that finished in the same divisional ranking (firsts play firsts, lasts play lasts.) within their conference which gives a total of 16 games.
It is only after playing all 16 games that the playoffs are seeded with the top seed playing the bottom, though that isn't always the worst team to make the playoffs.
nearly all pro sports have a season before determining who are the top teams.

In a tourney, that "season" would be the first couple days. So why not have the "top Players" go head to head early?
Just commenting that with a lot of the big tourneys, it seems that they avoid having this happen to build up "excitement" for the later games.

This all takes into assumption that 40k is inherently balanced. It is not, nor will it ever be. There will always be a top meta, and if a strong faction encounters a good player with a top meta list, no amount of skill will overcome that.

40k tournament scene has always been dominated by top meta chasers, you don't see GSC winning out GTs because they are widely viewed as the worst faction in the game. Tournaments are a bad indicator of faction balance.

How often players of various skill levels can achieve success without gimmicky or jank tactics would be a better indicator. A player with very little knowledge of the game can acieve success with most of if not all of the Astartes flavors, including Custodes. Now take a weak or untrained player and give them a Nid list, and they will not achieve success, or hardly any,

 skchsan wrote:
That's rigging the odds. Sports teams dont PURPOSELY pit top seed against bottom seed - it just happens that way because most sports are divided into leagues (there by creating sort of hierarchy that may not be reflective of the teams' ranking overall. Bottom ranking team from one league may be better than 1st seed of another league) that come together at the end in tournament format.

Not in most leagues.

The NBA has two conferences and when the playoffs start the teams are matched up based on where each team finished within their conference, with each conference's playoff bracket seeded 1 v 8, 2 v 7, 3 v 6, 4 v 5. As teams play teams from outside of their division and conference you tend to have an idea of who the strongest and weakest teams are.

The NHL used to use this format but has now switched to, 4 conferences. In general, they are seeded 1 v 4, 2 v 3 but there are occasional crossover teams due to the wildcard rule so one conference my be seeded 1 v 5, 2 v 3 and the other will be seeded 1 v 4, 2 v 3.

MLB uses a similar seeding system with the best team facing the winner of the wildcard game.

The NFL also uses a wildcard and then seeds top seeds versus bottom seeds in the playoffs.

 helgrenze wrote:
In a tourney, that "season" would be the first couple days. So why not have the "top Players" go head to head early?

You could, but it makes more sense to use results from previous tournaments to seed players as you want a tournament to shed the chaff as soon as possible so that you can get to the matches that actually matter. What good does it do to have a top player go out early and a play for fun goober advance just because of random seeding?

FezzikDaBullgryn wrote:
How often players of various skill levels can achieve success without gimmicky or jank tactics would be a better indicator. A player with very little knowledge of the game can acieve success with most of if not all of the Astartes flavors, including Custodes. Now take a weak or untrained player and give them a Nid list, and they will not achieve success, or hardly any,

This says more about the skill floor for each faction than it does about how powerful they are. Sometimes the easy faction will also be very powerful in the hands of a skilled player but other times you get a faction that has a high floor and low ceiling that can feel strong in casual play while doing nothing in tournaments.

I suppose I was talking more about balance there then actually being powerful, which is not the same. Point taken Canadian 5th.

You could, but it makes more sense to use results from previous tournaments to seed players as you want a tournament to shed the chaff as soon as possible so that you can get to the matches that actually matter. What good does it do to have a top player go out early and a play for fun goober advance just because of random seeding?

Over a 6 game first 2 days, top players are just as likely to still finish near the top. Having 1 loss would still affect the rankings, but it would be less likely to have a top 10 of undefeateds.
I am not saying that the top players should be auto matched with each other in the first couple games, but they should have the same chance to play a top player as everybody else.

So let's say that there are 60 people at the event. The first 6 games are true random pairings, with no repeats. Win/loss is the first thing to look at, then something like margin of victory, then alternate goal points.
For Faction rankings in the above, if there was an even distribution of 10 factions (6 each), if one faction averaged 5 wins (25 of 30) and another averaged 3 (15-15), that would be a fair indicator of the power rankings of those factions.
However, if 2 of the top 5 players played the same faction that finished with a mid level (3-3 average) record, that would skew the power rankings.

FezzikDaBullgryn wrote:
I suppose I was talking more about balance there then actually being powerful, which is not the same. Point taken Canadian 5th.

Balance isn't determined by what two totally new players do with any given pair of armies. Especially in 40k where things like which models each player has access to, what list each player builds, which strats they use and when, target priority and other factors can often determine as much as the exact faction played does. It's only at more advanced levels where players have larger collections and aren't making basic mistakes in target priority, movement, list building, etc. where you can start testing for balance. Ideally, you'd have matchmaking like online games have where players are matched by their skill level and it isn't hard to access the models you need for any given list, but unless GW goes all-in on translating 40k into a computer game we'll never get this.

 helgrenze wrote:
If they fall to a "for fun" player, were they really that good? Did they deserve to be preranked as top teir?

What? People are saying they should seed things so that top players face one another in round one while for fun players face each other round one, the top player going out would be because a match 4 style match happened round 1 not because they lost to a scrub. Have you even read what people are asking for in this thread?

 Canadian 5th wrote:

 helgrenze wrote:
If they fall to a "for fun" player, were they really that good? Did they deserve to be preranked as top teir?

What? People are saying they should seed things so that top players face one another in round one while for fun players face each other round one, the top player going out would be because a match 4 style match happened round 1 not because they lost to a scrub. Have you even read what people are asking for in this thread?

When looking at FACTION rankings, you have to look at the Average win/loss record. Top players likely will have similar lists if playing the same faction, and having more than one of them in the top finishers will have an affect on that, but basing a power level ranking on JUST those players skews the results.
See my edit above.

 helgrenze wrote:
Over a 6 game first 2 days, top players are just as likely to still finish near the top. Having 1 loss would still affect the rankings, but it would be less likely to have a top 10 of undefeateds.
I am not saying that the top players should be auto matched with each other in the first couple games, but they should have the same chance to play a top player as everybody else.

Why? Either this changes nothing and is thus pointless or it knocks out strong players early and leads to one-sided stomps in the later rounds as weaker players fluke into good records.

So let's say that there are 60 people at the event. The first 6 games are true random pairings, with no repeats. Win/loss is the first thing to look at, then something like margin of victory, then alternate goal points.
For Faction rankings in the above, if there was an even distribution of 10 factions (6 each), if one faction averaged 5 wins (25 of 30) and another averaged 3 (15-15), that would be a fair indicator of the power rankings of those factions.
However, if 2 of the top 5 players played the same faction that finished with a mid level (3-3 average) record, that would skew the power rankings.

What is this supposed to fix? Is it meant to make the tournament more 'fair' by making player's records more random? It doesn't ensure that top players see more list variety as they'd still end up mostly playing weaker players running lists full of garbage units. It doesn't help ensure the best players advance as it actually makes it more likely they get knock out early. So again, what is this supposed to fix?

 Canadian 5th wrote:

 helgrenze wrote:
Over a 6 game first 2 days, top players are just as likely to still finish near the top. Having 1 loss would still affect the rankings, but it would be less likely to have a top 10 of undefeateds.
I am not saying that the top players should be auto matched with each other in the first couple games, but they should have the same chance to play a top player as everybody else.

Why? Either this changes nothing and is thus pointless or it knocks out strong players early and leads to one-sided stomps in the later rounds as weaker players fluke into good records.

So let's say that there are 60 people at the event. The first 6 games are true random pairings, with no repeats. Win/loss is the first thing to look at, then something like margin of victory, then alternate goal points.
For Faction rankings in the above, if there was an even distribution of 10 factions (6 each), if one faction averaged 5 wins (25 of 30) and another averaged 3 (15-15), that would be a fair indicator of the power rankings of those factions.
However, if 2 of the top 5 players played the same faction that finished with a mid level (3-3 average) record, that would skew the power rankings.

What is this supposed to fix? Is it meant to make the tournament more 'fair' by making player's records more random? It doesn't ensure that top players see more list variety as they'd still end up mostly playing weaker players running lists full of garbage units. It doesn't help ensure the best players advance as it actually makes it more likely they get knock out early. So again, what is this supposed to fix?

Lets look at the Las Vegas open. (40k)
Among the top 10 there were 7 "Space Marines", 2 Aeldari, and 1 Ork. The top 2 were "Space Marines" meaning that one finished 9-0 and the other 8-1. Similar results would be expected in the other final round matches. Since Aeldari finished 3rd, their record is likely 9-0 as well, and their finish was decided on points.
If you based the "Power Level" on these results, "Space Marines" would look like they were nearly unbeatable.
But, how many people actually played "Space Marines" and what was the AVERAGE win/loss record?

 helgrenze wrote:
Lets look at the Las Vegas open. (40k)
Among the top 10 there were 7 "Space Marines", 2 Aeldari, and 1 Ork. The top 2 were "Space Marines" meaning that one finished 9-0 and the other 8-1. Similar results would be expected in the other final round matches. Since Aeldari finished 3rd, their record is likely 9-0 as well, and their finish was decided on points.
If you based the "Power Level" on these results, "Space Marines" would look like they were nearly unbeatable.
But, how many people actually played "Space Marines" and what was the AVERAGE win/loss record?

How does your suggestion change this?

As for your question, why not go and look up the tournament and see for yourself. The match results should all be public data.

Neither really.

Pure W/R has several obvious problems:
Most obviously, it counts play at all levels, when only the highest level of play is really relevant for balancing.
Second, high play factions will have winrates trending towards the mean [50%] due to mirror matches and general high play volume across skill levels.

On the other hand, only top finishes does separate out the bad lists and players, but has a small sample size easily skewed by player preference, and is also highly affected by the size of events.

As player skill can affect an armies performance, I would take out outliers and average the results. I would take the middle 50% of games so that the best players and the worst players don't distort the results. Or maybe a bigger chunk of the middle if there are enough results.

The entire game is based off of inherently random outcomes, it's impossible to factor "luck". If I perform statistically high in rolling all 6's, and my opponent rolls just 1s, I will throw the data. This has happened before at GTs, back in 8th when a freak ork list made it into top 8 witth an entirely SM meta remaining 7. It's not unheard of. Skill doesn't replace luck in these situations.

FezzikDaBullgryn wrote:
The entire game is based off of inherently random outcomes, it's impossible to factor "luck". If I perform statistically high in rolling all 6's, and my opponent rolls just 1s, I will throw the data. This has happened before at GTs, back in 8th when a freak ork list made it into top 8 witth an entirely SM meta remaining 7. It's not unheard of. Skill doesn't replace luck in these situations.

You don't roll 6s for five games in a row.

Example of the point I am trying to make.

A recent tournament had 65 players.
The winner played Iron Hands, a subfaction of Space Marines. His record was 5-0.
6 other players also played Iron Hands. One of those placed in the top 10. The AVERAGE record across the 7 players was 3-2. The worst had 1 win.
The rest of the top 10 were, in order, Nids, Blood Angels, "Imperium", Imperial Fists, "Imperium", Aeldari, Iron Hands, Chaos.
The bottom 5 were, Chaos, Stealer Cult, Chaos, "Imperium", Astra Militarum. Total wins among these was 2.
There were 16 Chaos players, Average record 2-3
There were 3 Nids players, if you count the GSC, Average record 2.5-2.5 One was the #2 with no losses, GSC had no wins, and the third managed a tie.
There were 5 "Imperium" players Average record 2.5-2.5

If you drop the outliers, top finish and bottom finish, most factions wind up about 2.5 wins on average. Also interesting to note, The players at the mid point of the standings, #32 and 33 were the ones that pulled the tie and finished with 2.5-2.5 records.

Factors that play in the differences include the lists, player skill, and luck. The list is the only one that plays any part in the power rank of a given faction. Skill is about the player's abilities and luck is random chance.

Top finishes don't tell you much about a faction.

A faction can never get it into the top of big events, and yet be a crushing experience to face at casual level. SM and knights have been good examples of this.

Spoletta wrote:
Top finishes don't tell you much about a faction.

A faction can never get it into the top of big events, and yet be a crushing experience to face at casual level. SM and knights have been good examples of this.

Knights and SM topped tonnes of events.

 vict0988 wrote:

Knights and SM topped tonnes of events.

yeah, unlike with 9th, 8th had the problem that an army with 2 BA captins and 15 scouts, was considered a BA list as long as the BA captin was a warlord . Pure knight armies weren't doing that well in 8th, even durning the reign of castellans.

I would say a mixture. Average W/L ratio across tournaments, number of top eight placements, and then some kind of weighting by representation.

It makes sense that any faction that is overwhelmingly represented will dominate placements by sheer volume, unless the faction is literally unplayable trash or the competition is only beatable in very niche circumstances like the deathstars of old in kill point missions.

 Daedalus81 wrote:

FezzikDaBullgryn wrote:
The entire game is based off of inherently random outcomes, it's impossible to factor "luck". If I perform statistically high in rolling all 6's, and my opponent rolls just 1s, I will throw the data. This has happened before at GTs, back in 8th when a freak ork list made it into top 8 witth an entirely SM meta remaining 7. It's not unheard of. Skill doesn't replace luck in these situations.

You don't roll 6s for five games in a row.

Well in theory...

I think "luck" matters because certain outcomes depend on a few dice. The biggest is who goes first. The second is "Do you make a charge?"
I've seen many winning tournament runs amount to saying yes for both - and ending abruptly in the first game they both came up "no".

But really "power" is about this being repeatable. You'd sort of expect every list to win a reasonably sized tournament *at some point* on the "I'll roll hot, my opponents will all be unlucky". I'm not sure it happens though - this may however be because good players don't tend to play bad factions.

9th doesn't seem that bad (probably because going first is a far bigger skew) - but when you have say the same tiny handful of lists winning tournament after tournament, it seems likely something is skewing the luck in their favour. If you rolled a lot of 1s while playing late 2019 Iron Hands, odds are the other guys at the Tournament playing them didn't.