D's tournament advice

[D_acolyte]

So here I was thinking I could just do something simple and not have it rattle around in my head for weeks.... I FAIL AT LIFE!!

What I am about to talk about may make you think I do not like tournaments or enjoy them. This is not true. I do like them and try to go to as many local ones as I can.

So in case you have not guessed, I said the last time I was going to do something on the differential..... well that escalated quickly to the tournaments in general. So a little about my history and view on tournaments, I went to my first tournament for Warhammer some 16 years ago at the age of 14 or there about and have always viewed them as a trial by fire. Do I win tournaments? No, or at least not often enough for me to be confident saying I do, I try always to have fun and a good showing.

So what I will cover is the differential, a little about list design from my opinion, and keeping your eyes on points. This is not new stuff but general good advice.

So some people who did not need to take regression may not have think a lot about differential. This is relatively easy, it is the difference between your observed and your predicted. So how does this effect the game, well if your predictive is that the players are evenly matched then it is simply the difference between your score and your enemies. So if you wanted to you could incorrectly say that if I have a better differential then I am the better player, do not do this one it is often a negative frame of mind and two there is so much involved in the game that this take a very simple outlook.

How are differentials used in Wyrd tournaments, they are the first go to tie breaker. Which works well for the most part.

So to recap, in tournament standing is it better to have a 10 to 9 games or a 9 to 7 game as the victor? It is better to have the 9 to 7 because it adds more to your differential then the 10 to 9.

Now you may be saying that simple stuff D, well your right. This is simple stuff, but what does it take to get this to happen. With GG16 having it so you can not declare any schemes it take guess work if you want to have a shut out, that is a 10 to 0 game. I know some people who think of it as if there are 20 points to grab instead of 10, which is not a bad way if you are looking at how to visualize stopping your enemy from gaining points. The schemes are varied enough and tricky enough that it is hard to stop a person from getting a few points.

The 10 10 way

So of the very many styles of play to come up around the tournaments the one I could never understand is the 10 10 theory. I would love for someone more voiced in this to please comment but I will give it a shot from what I understand. So the basic concept is that you will choose schemes that revolve around your opponent as little as possible. A good example is bodyguard on the mechanical rider and breakthrough. In doing so you are working to get your 10 points irrespective of your opponent but when two players which do this meet it will often end in a draw. So for starters, a draw is not worth as much as a win is normally and then even if you say “o shoot I need this to be a win” chances are you will only win by a slim margin from what I have seen. This interns really hamstring you in trying to be in the top 3.

Confrontation way

So this should be of no surprise by now but I am a confrontation base player. I tend to take 1 kill scheme and one other that I can do around the sides. As a result I almost never draw, I tend to win to varying degrees or loose big. I mean loosing big, I once had 2 wins and a crushing loss and ended up with a negative differential. Is this to say you should play this way, no I would not tell anyone that this is the way to go but it is useful to keep in your mind.

Building your list

So this can be really the meat of this but I will only talk about 2 cases. Building to keep the scheme choices open and building to counter the asymmetry of the faction.

So building to keep your scheme selection open is a pretty simple things. I do not care about my opponent schemes but if I see there list I might be able to narrow it down, for instance if there is 1 non master upgrade then they are probably not going for show of force. When I see show of force I tend to slap between 2 and 3 non master upgrades into my list. When I have public demonstration I put in 15 stones of minions, they may be the only minions but the goal is to keep my enemy unsure of what I took and to make a shell game of possibility. Otherwise field what you want.

So this next one in an interesting one. The factions are asymmetric in there styles and options, this is to say that the factions tend to be noticeably different in there approach/design in the game. For instance, the number of guns in Guild vs that in Neverborn. This leads to the feel of a faction as well as what your can sort of expect going opposite of it. A good way to psych a person out is to brake this conception. For instance, I will play a defensive force with Neverborn or try and dance around with some other times. I do not do this often in tournament because it is not my style but it is nice to do sometimes, I keep threatening that one day I will bring guns in my crew. Also just as playing in a way not normally associated with your faction in your meta is a good way to change things so it is with how people see how you play.

So long ago people would say my style is based around doing the unexpected or foolish moves, now they would say that it is about the blitz; the truth is it is all about pressure. Switching the overall style of your force is a good way to apply the pressure of uncertainty to people. Some people get worried when there plans have to change because the battle has shifted in an unforeseen way. If your one of those people play verse those that can do this to get better at adaptation and breath, try to come up with a viable strategy in 10 seconds or less.

Keep your eyes on your schemes and strategy.

So this goes sort of without saying, but you cannot win without points. To this extent 3/5 of the points possibly are in your schemes and where this realization is easy in the afternoon game in a tournament sometimes it seems like the hardest thing to remember. This is because it is a contest of mental endurance and focus especially by game 3. To compound this sometimes tournament organizers may have chosen some odd combination, like interception with search the ruin and show of force in the mix. Round 3 is the gauntlet, you showing there is often more important than any round in a tournament as one or more of the front runners are poised to loose and that might make them drop from 2^nd to 4^th or lower. I get it sometimes it is just nice to smash face and murder models, I really do get it as I feel the call to every game but no matter how seductive that call is resist it. If your goal is too place in the top 3 then keep your mind in the game and always assess your schemes.

So I feel all this could be much longer but that might be for another time. Please discus and tell me what you want next. Another post will becoming soon.

[tomjoad]

Reading this is really kickstarting my desire to write a long rant about how it is awful to use differential as the first tie breaker...I don't want to sound like sour grapes since I did get sort of hard done by diff at Califaux, but I have been pretty vocal for more than a year that diff is terrible at determining who had the better performance, so hopefully nobody will think that's my motivator.

The first tie breaker should always always bee opponent's game win %. This is superior even in a 3 round event, I think, and becomes more so the longer a tournament runs. Take this basic illustration:

Player A wins round one by 3 points, wins round two by 1 point, and loses round 3 by 5 points. 

Player B loses round one by 1 point, then wins round two by 8 points and round three by 6.

By differential, player B finishes far ahead (13 for B, -1 for A). However, strength of schedule will very likely show that player A had a harder road to achieve a 2-1 record, having played against undefeated players every round. By losing round one, player B has made it impossible for themselves to win the event, but it is now FAR more likely that they come in 2nd place than a player who doesn't lose until the finals.

There will always be a fair amount of luck in miniature games tournaments since we can only realistically fit 4-5 rounds maximum into a single day. Variance of matchups will play a large part in determining who has the best chances to win. I very strongly feel that abandoning diff and using a strength of schedule based approach would at least take a step towards rewarding the correct people for that variance.

[Lucidicide]

And how would you determine strength of schedule?

[valhallan42nd]

15 minutes ago, Aaron said:

And how would you determine strength of schedule?

That would almost require historical data based on the store's or a regional meta. I don't know if it's worth the paperwork.

[tomjoad]

14 minutes ago, Aaron said:

And how would you determine strength of schedule?

Opponent's record. This would be tough to calculate by hand, and that's maybe the best argument I've heard against it, but if Player A beats B, beats C and loses to D, and B goes 2-1, C goes 1-2, and D goes 3-0, then A's opponents would have a cumulative 6-3 record.

Player W loses to  X, beats Y, and beats Z. X goes 1-2, Y goes 0-3, and Z goes 1-2 as well. Cumulative opponent's record is 2-7.

Using only info from this event (no historical data needed), I would say that player A should finish above player W regardless of differential.

[D_acolyte]

3 minutes ago, valhallan42nd said:

That would almost require historical data based on the store's or a regional meta. I don't know if it's worth the paperwork.

For the record my only opinion on the posible change is I want this historic data, and it is probably not worth it.

[tomjoad]

Ultimately, I feel like, in an 8 person event, the two people who start out 2-0 should be very likely to finish in 1st and 2nd place. My proposal increases that likelihood, whereas the current system means that the person in 2nd is more likely than not to be somebody who lost their first round and then played against other people who were losing.

[Squirrelich]

My 2 cents, just for another data point: 3 round is hard. 3 rounds doesn't give a lot of wiggle room to create a good bell curve to place players on. There just isn't always enough placement/play data in 3 games to take out randomness and chance, like when top players face each other round 1.

I personally enjoy the differential being the tie breaker because it makes sense to me. It makes sense in that it shows how much you won your games by, and it makes sense because it forces you keep fighting hard after a win and even sometimes take what would be unnecessary risk if it were just a single game.

Background: new to Malifaux, experienced in WHFB tourney scene

[D_acolyte]

2 hours ago, 7thSquirrel said:

My 2 cents, just for another data point: 3 round is hard. 3 rounds doesn't give a lot of wiggle room to create a good bell curve to place players on. There just isn't always enough placement/play data in 3 games to take out randomness and chance, like when top players face each other round 1.

I personally enjoy the differential being the tie breaker because it makes sense to me. It makes sense in that it shows how much you won your games by, and it makes sense because it forces you keep fighting hard after a win and even sometimes take what would be unnecessary risk if it were just a single game.

Background: new to Malifaux, experienced in WHFB tourney scene

The win count will probably never be normal though you could do the it approximate normality at X observation but it is more a Poisson distribution, that is count data. Now if we had historical data on people there is a lot we could do, such as removing the randomness of who faces who first round and instead have the top players face each other first round, like Mason and anyone but me lol.

I agree that the differential is much easier to understand, has more utility and also allows for greater variance in who makes the top 3 which is a good thing in my book.

[Math Mathonwy]

On 7/2/2016 at 1:53 AM, tomjoad said:

Opponent's record. This would be tough to calculate by hand, and that's maybe the best argument I've heard against it, but if Player A beats B, beats C and loses to D, and B goes 2-1, C goes 1-2, and D goes 3-0, then A's opponents would have a cumulative 6-3 record.

Player W loses to  X, beats Y, and beats Z. X goes 1-2, Y goes 0-3, and Z goes 1-2 as well. Cumulative opponent's record is 2-7.

Using only info from this event (no historical data needed), I would say that player A should finish above player W regardless of differential.

You could simply use the total number of wins gotten by your opponents as the first tie-breaker. That isn't very difficult to calculate. And then differential as the second tie-breaker. So two people have two wins and one loss, the guy whose loss is against the guy who got three wins gets better position than the guy whose loss was against a fellow who only won one game.

[Squirrelich]

So hypothetical/anecdotal situation: players 1&2 both win their first two games and then tie game three against each other, so they go to tie breaker.

Player 1 played two lower skilled players to get to this point, he went 10-0 against both and their win/loss records were both 1/2. 

Player 2 played two higher skilled players, but not as good as him or player 1. He beat both of then 6-5 and their overall win/loss were each 2/1.

 

So, do you:

A ) give overall victory to player 1? This in effect rewarded him for soundly trouncing his opponents, though they were newer to the game. Meanwhile, this punishes player 2 because his games were less decisive, probably because his opponents he got matched against were more experienced.

B ) give overall victory to player 2? This rewards him for beating the tougher opponents. But this also punishes player 1 because although he played perfectly and had complete domination of his games, it is because chance dictated he play weaker players that he was able to do so.

(There probably are not right and wrong answers to this question since chance decided who played the stronger and who played weaker opponents and both players were equal)

Personally I'd give victory to player 1. Because imagine how bad it would feel to after 10-0 winning two games you get knocked to second by the guy you tied because fate dictated that those people would proceed to lose the rest of their games, something you had no control over.

Though full disclosure, I've always been partial to battle point systems that ditch w/l/d records entirely.

[tomjoad]

10 hours ago, 7thSquirrel said:

 

Because imagine how bad it would feel to after 10-0 winning two games you get knocked to second by the guy you tied because fate dictated that those people would proceed to lose the rest of their games, something you had no control over.

Imagine instead beating two of the best players in the room and tieing with a third, then being told that you get 2nd place because somebody else got to play a couple new guys and step on their necks, which is also something you had no control over. To me, there is no question that the player who faced the stronger competition on the day should be the winner here.

[Adran]

On 02/07/2016 at 0:27 AM, tomjoad said:

Ultimately, I feel like, in an 8 person event, the two people who start out 2-0 should be very likely to finish in 1st and 2nd place. My proposal increases that likelihood, whereas the current system means that the person in 2nd is more likely than not to be somebody who lost their first round and then played against other people who were losing.

The biggest downside to this is what if the top 2 players meet in the first round. the loser of this match now can't really get better than 3rd, even if they play better than the person that loses to the winner in the final. By making it diofferential you are stuill giving eevryone somethign to play for in all their games. 

I've never played a strength of schedule tournement, so I don't know how much this would change things, but I can imagine that it would make pairings much harder to organise, as you are having to calculate strength of schedules all the way through the event to determine the draw. 

2 hours ago, tomjoad said:

Imagine instead beating two of the best players in the room and timing with a third, then being told that you get 2nd place because somebody else got to play a couple new guys and step on their necks, which is also something you had no control over. To me, there is no question that the player who faced the stronger competition on the day should be the winner here.

The likelyhood that the second person you face is a complete walkover is not so high as they will have had the second best win in the first round. Meanwhile "apparent 3rd best player" only managed a narrow win in the first round as well, so you are looking at a fairly strange set up for this to happen.  (its certainly possible, but not that likely). 

 

I wouldn't choose to avoid an event based on it being decided on strength of schedule, but it does seem more likely the case that if the final is a tie, you end up with the decider being what the outcome of the losing "semifinalists" is. Which seems a bit strange. 

It places the first tiebreaker to be something that is out of the players control. VP diff might not be fair, but it is at least something the player can affect. 

 

[Squirrelich]

11 hours ago, tomjoad said:

Imagine instead beating two of the best players in the room and tieing with a third, then being told that you get 2nd place because somebody else got to play a couple new guys and step on their necks, which is also something you had no control over. To me, there is no question that the player who faced the stronger competition on the day should be the winner here.

I don't remember a time this has occurred for me so I can't objectively say I'd be alright with it, but the old local WHFB group already openly acknowledged that tourney wins come from match up schedule because of the rock/paper/scissor nature of the game. Everyone just kind of shrugged when you got a bad match up, even when you got the stronger opponent and got knocked out of contention because you got a smaller win margin the talk was just about how you got the more enjoyable game as a result. Maybe that background is tainting my opinion because this doesn't seem that horrible to me and my mind just says you have to do the best to win, not you have to do the best for the circumstances you were given to win. *shrug, different people, different experiences, different opinion. Maybe someday we'll meet at an event an resolve this in a civilized way, in a game of Malifaux

[tomjoad]

20 hours ago, 7thSquirrel said:

...you have to do the best to win...

I agree. I just think that doing the best means beating the best, not beating the tar out of a newbie. We agree that there's a luck-of-the-draw aspect to this, it's just a matter of who's luck should be rewarded.

[Squirrelich]

36 minutes ago, tomjoad said:

I agree. I just think that doing the best means beating the best, not beating the tar out of a newbie. We agree that there's a luck-of-the-draw aspect to this, it's just a matter of who's luck should be rewarded.

I didn't say it was the most impressive feat of the tourney, your tough wins should probably be worth more pride. But by the game's own definition of victory the larger margin winner won harder... more hard... whatever the correct grammar is

[tomjoad]

6 minutes ago, 7thSquirrel said:

I didn't say it was the most impressive feat of the tourney. But by the game's own definition of victory they won harder... more hard... whatever the correct grammar is

Well, the regular game just defines winning as having more points. There's no distinction for winning big in the rule book. Wyrd aren't constrained by anything when making the tournament rules. Shouldn't the "most impressive feat" be what gets celebrated?

[Squirrelich]

31 minutes ago, tomjoad said:

Well, the regular game just defines winning as having more points. There's no distinction for winning big in the rule book. Wyrd aren't constrained by anything when making the tournament rules. Shouldn't the "most impressive feat" be what gets celebrated?

Touche

What if the most impressive feat is the new guy only slightly losing to the best instead of getting stomped as expected? Becomes very hard to quantify. What if the guy who stomped the noobs was the best, but didn't get a chance to show it because he was matched with noobs? Therefore he loses to the second best who was playing closer matches.

But winning by more getting more points? That's easy to count not only who won, but between who won who did it by a larger margin. Makes things clear, even to an outsider who wouldn't know who is supposed to be the best in the room and who is just flipping jokers at the right or wrong time.

Again, each round after 3 round the issue gets less relevant as things should be getting more clear.

[Tawg]

1 hour ago, 7thSquirrel said:

Touche

What if the most impressive feat is the new guy only slightly losing to the best instead of getting stomped as expected? Becomes very hard to quantify. What if the guy who stomped the noobs was the best, but didn't get a chance to show it because he was matched with noobs? Therefore he loses to the second best who was playing closer matches.

But winning by more getting more points? That's easy to count not only who won, but between who won who did it by a larger margin. Makes things clear, even to an outsider who wouldn't know who is supposed to be the best in the room and who is just flipping jokers at the right or wrong time.

Again, each round after 3 round the issue gets less relevant as things should be getting more clear.

Secondary note regarding this whole topic (I think this quote might be slightly irrelevant, whatever):

 

While it's not entirely something you can do, especially at events which draw in larger crowds (Like Cons or whatever, maybe some local scenes draw in larger player pools), so maybe you can't help everyonnneeee, but!  If Differential is the goal, and this is something occurring at a local level, like your local meta at a FLGS, then doesn't Differential do the best thing for the game that can be said?

This system should, in theory, be the most encouraging to help all players to learn to play their best.  Actually, almost like @necroon's blog post here on the site, about wanting to be a better player.  He is essentially uplifting, not only himself, but the larger group of players he frequents with, attempting to raise their skill level and probably moral and or willingness to play at higher levels of play (See: Tournaments).  But the follow up to this is that, the less "noobs" there are to encounter (Say you helped player one, who could never score points on Turf War before), then you're decreasing your opponents' chances of luck-sacing into "insane differentials".

Obviously that only goes so far, but in theory, it means everyone should look to help each other, as a gaming community, in order to best hinder their future opponents.  But the scale of the event alters this outcome, and there will always be newer players or what have you.  But ultimately you could help curb it slightly, assuming you don't exclusively play with strangers.  Or maybe you only play tournaments, so you can't really impact the local meta by helping "train" people essentially.  There are plenty of things that could go wrong.

 

Personally, I'm not a tournament player, for most any game.  I don't particularly find competition compelling, and often find games lead to very narrow avenues by which such "competitive list" can exist (Which kills it for me).  40k is a fairly bad offender in this regard, IMO, because often there are "Correct list" to take, and deviation is not typically rewarded in any way.

Although I don't have a large enough local meta to play Malifaux (Without an hour+ drive), but I will say of all the games I play, I feel Malifaux has the most likely chance to not punish players for having a list that is not "The correct list", even if rather obvious good choices are present.  Granted you could also completely fumble list creation in Malifaux and have one hell of an uphill battle, but ya know.

 

Anyways, just a counterpoint, that this is potentially a boon to the community aspect of the game, in disguise.

[tomjoad]

I think that the current system encourages am experienced player to brutally pound a new player into dust if they're lucky enough to get such a matchup, because getting a 9 or 10 point win is so valuable under differential. Abandoning diff as the first tie break will encourage experienced players to instead offer advice and mentorship to the newb, since getting that huge win will be severely devalued.

We all agree that an advanced player SHOULD help the newb when possible, right? That helps grow community more than thrashing a 15 year old playing his or her first miniature game, after all. Why not adopt a scoring system that eliminates the disincentive to that kind of outreach?

[Adran]

I was thinking last night, and came up with a case that feels completely wrong on the strength of schedule (based only on the days tournament) 

14 player event, with 3 rounds. 

Player 1 wins the first game 10 -0. Opponent 1 loses their 2 remaining games. (Strength of schedule 0 wins 3 loses)

Player 1 then wins round 2 10-0, opponent 2 loses the last round (strength of schedule 1 win 2 loses)

Player 1 wins round 3 10-0. (strength of schedule for opponent is 2-1).

 

Player 2 wins round 1  2-1 Opponent goes on to win their rounds 2 and 3 (Strength of schedule 2 wins, 1 lose)

Plays round 2 wins 2-1. Opponent wins last round as well (SOS 2 wins 1 lose)

Plays round 3 wins 2-1 (SOS 2-1)

Final results 

Player 1 wins 3 games, scored 30 VP, concedes 0 vp. Differential +30, SOS (based on opponents results) 3 wins 6 loses

Player 2 wins 3 games, scores 6 VP, concedes 3 vp. Differential +3, SOS 6 wins 3 loses. 

Which one should count as the winner? 

 

I know that when I face opponents I will help them out. Occasionally this will cost me games. But most of the time it doesn't, and I might choose the time to give advice as just after it happened on some things (especially if the game feels close). And the score at the end doesn't always represent how close the game felt. (I'll admit I will try and play for a game that feels close, but that I win by a long shot against a newer player, if I think I am that much better than them). 

Prehaps points scored could be a better tie break, but it certainly has its own faults.  

[Tawg]

1 hour ago, tomjoad said:

We all agree that an advanced player SHOULD help the newb when possible, right? That helps grow community more than thrashing a 15 year old playing his or her first miniature game, after all. Why not adopt a scoring system that eliminates the disincentive to that kind of outreach?

Well in the middle of a tournament, nobody is going to be holding your hand.  Certainly not your opponent, at the very least.  There is a game to play, and if you come to a tournament, there is an expected level of proficiency.  Besides the fact that no one is going to coach a player on how to beat them during a match.  And outside help is obviously not something most players would stand for, because no one wants to lose because someone, not their opponent, was just tossing around tricks or helpful guidance.

So let's remove the possibility of either of our mentalities helping the situation in the present time of a good player vs a bad player during a tournament.  Because neither of our thoughts are going to lead to that outcome.

 

So what we expect then, is that away from the tournament, players aught to help each other.  I believe this gives incentive to players to help their weaker opponents in their control (e.g. in their meta, people they can coach in casual matches, possibly regularly).  And you clearly believe this should be a goal as well.

But your statement is implying that with this system, good players will exclusively choose to crush new players, bad players, and people just trying to have fun, because it's the second highest goal to winning itself in a tournament setting.  Regardless of if they are playing said tournament, or casually at a FLGS or with friends.  If that is the case, we have a different problem, namely a toxic community.

So like I said, in the circumstance that a poor player has entered a tournament and goes against a player in the know, then they will crush them no questions asked, whether differentials is a goal or not.  And new players will lose matches, and it will look brutal, until you learn to digest the information of what went wrong, and what could have been done different.  But suggesting that eliminating differential will stop this, in the heat of a tournament, is not possible.