Matchmaking system in Drafts

[Curious Minds]

Is there any matchmaking system in Drafts?
First ~130-150 games I had stable 80+% winrate with only 4/15 runs that was not victories, when even 3 of loses was [6-3] runs (11/15 full run victories with 3 clear [7-0])

After them, my winrate drops drastically:
Last ~100 games I had stable <60% winrate with 10/12 loses and 0 clear (7-0) victories at all.
Some losed runs now have a pathetic 1-3 score

Looks like there is hidden rank or any other matchmaking system depending on player skill in draft, am I right? Players now don't make mistakes at all and have top notch decks

Stats collected from big data (200+ games) so we can't say it happened by accident

[Anarchyの海豚]

I am sure there is elo. After winning a few game, my draw will get worse and worse to make me lose the game. I am not sure does the elo have relationship with how much money you spend on the game, sometimes on the events, when i match with a america with full of special card, even my build is good for the match, but i will get worse draw and lost the game, this usually happen in the second match in an event.

[Lucky_Star_Fan]

Draft is so much less stable than constructed. Did you keep track of how good you thought your deck was? What the nation matchup was? When I go 3-3 with what I consider a mediocre or worse deck I kind of expect it. If I lose to someone with a crazy deck I'm not surprised.

[Curious Minds]

Publicado originalmente por Lucky_Star_Fan:

Draft is so much less stable than constructed. Did you keep track of how good you thought your deck was? What the nation matchup was? When I go 3-3 with what I consider a mediocre or worse deck I kind of expect it. If I lose to someone with a crazy deck I'm not surprised.

I emphasized many times that there was huge data, and that that was stable.
130 matches of solid succes in a row (one more time - stable), than, INSTANTLY 10/12 loses in a row (not matches, global draft runs), and even 2 victories was 7/2
Yep, I tracked decks too, here it is if you are interesed (2 right columns are wins/loses in a draft series)
https://i.ibb.co/zVgcD0fm/1212.png

[Curious Minds]

Publicado originalmente por Anarchyの海豚:

I am sure there is elo. After winning a few game, my draw will get worse and worse to make me lose the game. I am not sure does the elo have relationship with how much money you spend on the game, sometimes on the events, when i match with a america with full of special card, even my build is good for the match, but i will get worse draw and lost the game, this usually happen in the second match in an event.

thanks!

[puschit]

You'll get paired against opponents that have an equal win/loss ratio FOR THIS RUN. With Elo as tiebreaker. However, only few players do play draft, which means, in most cases, the system won't be able to find suitable opponents. It gets less and less acurate the more wins you have above three because there simply won't be enough players left to match you.

10 out of 12 losses while doing solid for 130 games before sounds a tad bit odd but it is still within the realm of probability (especially since you didN't define what "stable" actually means for you). It could mean for example that you got 3 or 4 bad drafts ins a row (which WILL happen at some point sooner or later) and then start doing mistakes because you worried too much about it. In gaming circles this is known as "tilting". You experience a streak that is unusual compared to your normal performance and it frusrtates you so much that you start playing sloppily, which causes you to lose more.

[Curious Minds]

Publicado originalmente por puschit:

You'll get paired against opponents that have an equal win/loss ratio FOR THIS RUN. With Elo as tiebreaker. However, only few players do play draft, which means, in most cases, the system won't be able to find suitable opponents. It gets less and less acurate the more wins you have above three because there simply won't be enough players left to match you.

10 out of 12 losses while doing solid for 130 games before sounds a tad bit odd but it is still within the realm of probability (especially since you didN't define what "stable" actually means for you). It could mean for example that you got 3 or 4 bad drafts ins a row (which WILL happen at some point sooner or later) and then start doing mistakes because you worried too much about it. In gaming circles this is known as "tilting". You experience a streak that is unusual compared to your normal performance and it frusrtates you so much that you start playing sloppily, which causes you to lose more.

I understand all your takes but I'm not tilting as I can analyze myself. Clearly noticed that after the change moment opponents starts to play way too better with better decks with strong 1-2-3-4 drops one by one from opponent even at the first game in a series

Maybe, yes, It is just coincidence

[puschit]

It usually is. Haven't played draft in a while, will do some runs in the coming days.
Do you play frequently? What kind of time period are we talking where you did those 130 games? Was there a rotation in between or new expansion?
You said your results were very stable, could you be more specific, does that mean 3-3 most of the time? Because to my experience results are always wildly different dependent on your draft luck. I am doing well overall because experience is still worth more than luck in the lomg run, but I get all kinds of results, I do have enough 7-X runs to call myself a good draft player but I still drop out with a single or no win occasionally. I wouldn't call that "stable". And with such a huge amplitude I thinks 8 out of 10 bad runs are within the realm of possibility that will happen eventually - at least compared to my results.

I'll report back in a week or so. Maybe we should start wirting down results for science :)

[Curious Minds]

Publicado originalmente por puschit:

It usually is. Haven't played draft in a while, will do some runs in the coming days.
Do you play frequently? What kind of time period are we talking where you did those 130 games? Was there a rotation in between or new expansion?
You said your results were very stable, could you be more specific, does that mean 3-3 most of the time? Because to my experience results are always wildly different dependent on your draft luck. I am doing well overall because experience is still worth more than luck in the lomg run, but I get all kinds of results, I do have enough 7-X runs to call myself a good draft player but I still drop out with a single or no win occasionally. I wouldn't call that "stable". And with such a huge amplitude I thinks 8 out of 10 bad runs are within the realm of possibility that will happen eventually - at least compared to my results.

I'll report back in a week or so. Maybe we should start wirting down results for science :)

Thanks man!
All this games are from last Blood n Iron expansion time period

About stability - there is an image that I've posted in some posts above, you can check it out
https://i.ibb.co/zVgcD0fm/1212.png
(2 right columns are wins/loses in a draft series and by stablity I mean first 14/15 runs was 7 or 6 wins)
There you can see Instant drop of winrate after 7/2 USA Fin run

[puschit]

Only played once yesterday and lost hard, but I had a hangover and made many mistakes, I don't think that counts :P
Great that you already protocol your games, more players should do that, especially those that claim the matchmaking is rigging them. Anyway, you said 130 games, sadly we see only a small part of that. I can see the harsh drop of course, but according to my memory my results are similar to what you experienced in the second half: High volatility.
What you should also note down is the day and maybe even time you play. For example you'll always get easier opponents on Saturdays and Sundays where lots of dads and casuals play, players with less experience, some of them maybe using a free ticket they got somewhere. So, if you used to play on weekends and then played the last dozen games midweek, that might explain the difference.
There might also be an influx of better players. Maybe it's just the mobile crowd that began playing months ago and are now catching up.
Please keep on writing results, I'll do the same, this time sober :)

[Curious Minds]

Publicado originalmente por puschit:

Only played once yesterday and lost hard, but I had a hangover and made many mistakes, I don't think that counts :P
Great that you already protocol your games, more players should do that, especially those that claim the matchmaking is rigging them. Anyway, you said 130 games, sadly we see only a small part of that. I can see the harsh drop of course, but according to my memory my results are similar to what you experienced in the second half: High volatility.
What you should also note down is the day and maybe even time you play. For example you'll always get easier opponents on Saturdays and Sundays where lots of dads and casuals play, players with less experience, some of them maybe using a free ticket they got somewhere. So, if you used to play on weekends and then played the last dozen games midweek, that might explain the difference.
There might also be an influx of better players. Maybe it's just the mobile crowd that began playing months ago and are now catching up.
Please keep on writing results, I'll do the same, this time sober :)

Ok thanks! =D

"you said 130 games, sadly we see only a small part of that" - we see not 130 but 204 games there. I mean there is 204 games on screen that I've posted. By games I mean matches, not full runs
The idea with casuals on weekends is indeed interesting, but I was playing every day in free random timings, so that was not that problem

[puschit]

My run yesterday was a mixed 3/3. Had a US/GER deck that was good lategame but a bit too slow.
Maybe it was the other way and you got overly lucky in the first runs?

The thing is: The cards you get offered are NOT totally random as evidenced that you usually get an elite card offered last, so there seem to be brackets of some kind. Nobody knows which. And you'll notice that some cards are offered more often than others. And that there are always some cards that get offered again and again. As if you are pushed to play a certain deck/style/combination. I don't think there is an actual agenda but but maybe they want to see certain cards being played, maybe they try to balance this way. I doubt that because US has always been too good for so many years now, IF they are trying to balance it certainly doesn't work :P
However, it means that you don't really get cards offered according to what you'd expect when cards are offered at random from the entire pool of cards. If it was truly random we would a) get a broader selection and b) get cards proportionally to the frequency of cards of certain mechanis/characteristics. But that just isn't the case. For example if you pick Japan you'd expect to get offered a lot of cheap, fast suicidal type of units but you just don't.

And that is problematic when drafting because going for syngergies is a huge gamble and you can't use stochastics. For example 2 weeks ago I pickes US/USSR and got a few cards offered early on that would fetch and buff covert cards. Since I still had 25 picks to go and since US and USSR do have the most covert cards I thought it would be ok to go for it. But I didn't get offered a single covert unit for the remainder of the draft ...
That's also one of the reasons why Japan is such a bad pick in draft: It has so many mechanics like self-kredit-denial, self-discard, specific bombers that only work in conjunction etc. that it's way too hard to pick something decent that synergizes. And you can't even count on getting some quailty weenies even though they are numerous enough.

So, while drafting, you are better off sticking to solid cards that always do their thing, drafting a good mana curve, looking for enough removal and card draw etc. Which leads to the nations that are usually picked. However, given the nature of the drafts it's still tremendous volatility for me. So you pick US main which is good and you pick some heavy hitters because you know you'll get mana ramp eventually ... only that sometimes you don't. You might notice early that you'll need more cheap units but you might get offered only expensive ones or orders 10 picks straight. And then there are drafts where every second pick is a no-brainer. For me, it has always been volatile, especially since Brothers in Arms. Playing and picking well does make a difference of course but there will always be a huge amount of runs that just won't go your way no matter how smart you are. At least that's my experience over the course of many years.
So, maybe the exceptional early win streak was the anomaly, not the second half. And I mean ... mathematically the global average wins is 3. Which means if you had 6 and 7 wins consistently forever this would mean LOTS of other players would have to drop out with 0, 1 or 2 wins to give you that high a score. And guess what, players that lose too often in a format where they have to pay an entry fee will stop playing at some point. You need 3 wins just to get your investment back. If you score lower than that regularly it means you are burning gold. And that's on top of the beating your ego got.
Therefore, draft is usually a pool of sharks unless you run into people using free tickets.

[Curious Minds]

I agree with this. I understand that Japan have mostly aggro or combo cards, and both are bad in draft, as you can see I had 0 Japan runs and 0 Japan 2nd nation runs.

Yep, me too noticed that some cards highly suggested in some drafts, if system decided that you need Supply Chain this run - declines will not help. You will get constant repeats of some trash draft cards if you ignore them and it clogs a lot

Global average wins is 3, but I had 80%+ winrate in first 120 games so I guess this statement that 3 is average and it's ok that I droped in winrate is not correct because it's all skill depended. Every player have his "normal baseline" of winrates. It's not ok if Jking will have 3/3 average. I guess you too have much more than 50% winrate in drafts and it's not ok if some day you will suddenly drop to 50% winrate in long term