I feel like H2H is such a terrible way to decide a tie breaker especially as a first tie breaker. SEN have better stats than KRU on EVERY STAT and has almost a +50 round differential and KRU only has a 14 round diff I feel like map win loss and round diff should 100% be the main tie breakers cause having H2H just makes every single stat useless
0
H2H is so dumb
posted in
General Discussion
Comments:
Threaded
Linear
Fair enough, Let me give you a proper reason.
Round difference and map difference is a silly metric to take into account, because it doesn't prove whether a team is better than the other with the same points. Round/map difference would make sense if every fight in Valorant was transitive but that is never the case. If Kru have beaten Sen, then it makes no sense for them to be below them. H2H is even more impactful in deciding the 4th seed. Lets assume 2g beat SEN and Loud win tomorrow. Even if 2G has a better map diff, EG should still qualify because they have beaten them.
Like it or not, Valorant rankings have to be reactionary to make sure the best team in that moment is the one that qualifies, not the team with consistent results. So for now Sen is a weaker team than Kru.
if 2G have a better map diff and round diff then no 2G should definitely get the tiebreaker over EG because they showed that even after having such a HORRENDOUS early few games they still managed to overall do better against the same teams that EG played against.
It definitely becomes a controversial topic when it comes to lower seeds but thats how you make every group stage game count, or else it ends before youve even played 5 games sometimes before youve even played 4
While yes not every gun fight is equal, comparatively to the amount of times you play each other in a group stage, the average becomes far more controllable when using rounds and maps to decide who was better than who. The average gunfight will more or less play the same through every game whereas the same matchup could be vastly different with each play due to teams just not always being in form.
There are more rounds and maps played than there are H2Hs, its a much safer metric to use to decide the better team
A fight for the higher seeding is the same as a final. It doesn't matter who looked better against the weaker teams. It depends on who beat the other team. Here, Kru have beaten Sen. Hence, in this moment of time, KRU are better than Sen. Hence they get first seed.
If according to you, every group stage game should count, then I'd argue that the most important game was the one between Kru and Sen and the results of the other games get low priority.
at this point in time we dont know who is better than who because they faced each other 4 weeks ago and both have shifted in form multiple times over that course of time.
Im saying map diff and round diff are a much more averaged and normalized metric to score teams to see who did better than who. There a reason every ohter league of sports doesnt decide tiebreakers through H2H
And Head to Head isnt a good metric to know wish team is better, because you have beat some team that your opponent hasnt beat, how the fact EG didnt had beat 100T, make Sentinel a worse team than KRU ?
A win is a win, beating your direct opponent is already a huge game changer, even without this rule, they are no reason to add it. Mathematically speaking, Head to Head is the worst way to sort out team in a round-robin format, the concept of round-robin is to give 1 point to any team that beat another, beating your direct opponent is the equivalent of 2 point (You win 1 point and you deny one), with head to head it's the value of 2.5 point, wish is insane (especially if the game was played on the start of the split). It sounds fair on the paper, and it's easy to believe why this is fair, but when you do the math, it doesn't. (Reverse head to head is bad, but it still may be better, but it wouldn't be accepted by people anyway) It can also lead to ironical situation where a team that start the split by loosing against a mid-team and end the split winning against the best team would be considered as worse because they get beaten by them 2 month ago.
Head to head is not only bad for competitive integrity, it's also bad for viewer experience, for 2 reason : it kill the suspensen (With other way to break tie, you have more game that have stakes), in some situation the math become too hard for most of the people
Mapdiff is the best way to sort out team (Wouldnt help in that situation), then map win percentage is the 2nd best way to sort out team (still not helping), both doesnt have the problem of roundiff you point at. Then it's pretty open, you can use round diff wish is a pretty easy metric to analysis for the viewer (Would make SEN 1st), you can use Buchholz per match win wish is calculated by doing the sum of point of every opponent you has beat and can be also visual for the viewer (This one would favor KRU), of you can even make a BO1 decider so we can have fun in high stake game without breaking the schedule (This would be the more fair and favor none)
I don't care about Sen, they are rule that can be used that would still make KRU first, and I agree roundiff has some issues, but head to head first is, indeed bad
the second match of a group stage isnt push comes to shove tho is what im saying. that match was no grand finals, there wasnt much weight behind it even if it meant seeding, why not just be normal like every other sports and actually utilize the stats to see who was overall better throughout the groups?
By your own argument, EG winning against Kru wasn't a Grand Final either. So we can just pretend that Kru didn't lose a game.
The reason this is not like real sports is because the difference between a bad valorant/esports player and a great one is very small. Whereas, in say Basketball, the difference is humongous. So Valorant rankings have no choice but to be reactionary.
At the end of the day, SEN lost to Kru, and Kru were the best team in the group stage.
thats an INSANE strawman cause i never said to just throw the game away completely. That same game brings with it those same round stats and map stats that show the overall performance of a team.
No the difference between good teams and bad teams in valorant are huge my guy, put a bottom ranking chinese team into any other league and theyd get stomped, same goes for KOI, same goes for Giants and Apeks and a ton of other teams. Even if that were the case i dont see the reason why H2H then becomes anymore viable at all.
You say KRU were the best team in the group but there is no metric to measure that, even by my proposal to use rounds and maps ti doesnt exactly tell you that sen were the best either, it just gives you a better idea of who performed the best throughout the weeks against the same teams ON AVERAGE
Makes complete sense to me, honestly. Team A and Team B are at the same level, but Team A beat Team B so they get priority seeding. Yeah, it can be tough, but any tie breaker's gonna be tough for whoever's negatively impacted, and frankly this method seems the most fair to me.
(Also the other stats do matter because if there was a 3-way-tie they'd be used to decide who gets priority seeding)
yea but thats such a rare occurence that the stats are most of the time useless.
Why does any team care about winning 2-0 why do they care about getting huge round diffs if its just gonna be useless at the end of the day? I dont watch much esports anymore or regular sports for that matter but i feel like the norm was definitely to look at stats before H2H in almost everything i watched
to add onto the above example:
Now assume Team A and B are the exact same again. This time there is a Team C that is competitive against A and B, but wont make playoffs.
Both Team A and Team B will end the season on a tie (we disregard who won the h2h matchup and only look at final "score").
Team A has to play Team C early into the season. Since team C is competitive it is a close game and they barely gain any rounds/maps.
Now by the time Team B plays Team C, Team C is already eliminated from playoffs and is making early roster changes to prepare for next split.
Team B crushes the newly formed Team C and gains massive round/map differential putting them ahead of team A narrowly (works the other way around too where Team C gets better post roster change).
Team B now wins the tiebreaker because of external factors neither team had a chance to influence. The tiebreaker is not decided by Team A, but by a 3rd team.
Thats why h2h is generally preferable, since it isolates the performance of the teams involved in the tiebreaker.
(This isnt to say there arent any scenarios that feel bad with h2h either, but I hope cloudberry's and my example illustrated why its good)
no you definitely have a point here but my whole point is that when were talking of qualifying through map win/loss and round diff most teams wont be out by week 5 unless they just go 0-5 or 1-4, in which case they were trash the whole stage anyway.
Also its not like round diff would be the primary metric, it would be map wins which wouldnt be a farmable thing just because a team is bad. You will at the very most win a +2 map diff regardless if youre opponent is good or bad so the analogy doesnt really work here
seeding and trophies are 2 completely different things though. The point of a group stage is to figure out who was the best team overall not to find out who is the best team right now.
SEN were weaker than KRU for their matchup but destroyed every other team that KRU had pretty close matches with. At the end of the day objectively sen did better but because KRU were better than them on week 2 they get the 1st seed.
Either way tho regardless of better this better that my main point is just that it makes the stats completely useless and that H2H has never even been a clear show of who is better. Just like TG beat Optic in stage 1 2022 yet they were terrible when they rematched again, we all know that 1 win doesnt mean anything in valorant
but thats exactly what im saying tho, using round diff and map count is a much more decisive way of telling who was overall better througout the group. It would take multiple teams playing worse than other weeks to be able to best the rest of the group which is a much less luck based metric than H2H cause val is just a game thats gonna have teams be inconsistent.
Loud and Optic were the most consistent team and they had a ton of losses during groups and regionals (maybe not so much loud)
but it doesnt and we know this already, time after time we see teams beating those who beat them before, Optic v Loud (in almost every rematch), Optic v TG, FPX v DRX, FNC vs FPX, SEN v GenG, T1 vs EDG, PRX v EG, its literally engraved in the history of val that winning today doesnt mean youll be the better team tomorrow
why is it not used in most other sports then if that were the case? Most leagues make points because they understand that head to head and "goal diff" isnt enough to show who the better teams are and yet here we are 5 years into val still using such a terrible metric that weve seen time and time again mean absolutely nothing even at the international playoffs level
Discord