Monday, November 22, 2021

Competitive Cribbage - Winter 2022 Season

Welcome to the Winter 2022 Season of Competitive Matchmaking for Cribbage Pro! Yes, it is still 2021 when we start this season, but I think we have had enough of 2021. The Autumn 2021 season has ended, and in a way that I honestly did not expect. So many new faces on the Top 50 leaderboard this last season, several of which came in at the "last minute" to make their way into the top rankings, and of course a new 1st place player as well! Please continue to tell your friends about competitive cribbage, and of course the chance to win the prizes! Thank you to all who were once again able to participate in this great game of cribbage!

As many will attest, it takes some significant commitment to achieve this, and so I want to very sincerely recognize that hard work here. I'm humbled by the individual effort put in by so many. That said, I also hear that it would be nice to lower that level of work a bit to allow more people a chance to participate. This season we are making some smaller changes, but won't be making the bigger changes until next season when I do plan to include more of your suggestions. Before I get into the more minor changes this season later in this post, let me take a moment to recognize the top players last season.

In 1st place, we have "ThreeG". Great job on achieving this rank and beating out some great competition! 2nd place is "iPeg", and I am sure you are all deeply saddened to see them fall from 1st place, after sitting there multiple seasons. In 3rd we have "Dhrun", which again I know took a lot of work to get there! There were several big improvements again this season, but we really saw a lot of new players in the Top 50, so congratulations to you all as well.

As with past seasons, everyone who finished the season in the Top 50 ("Recent" list), have been awarded both a special in-game "board peg" as well as Cribbage Pro Gold that can be used in the Cribbage Pro Contests system and then redeemed for cash (awards must be used at least once in a Contest to be cashed out to USDC, see the full terms and conditions for details). The top finisher is awarded the "crown" board peg, and all others in the Top 50 are awarded a "star" board peg. These pegs are shown to everyone when playing in online multiplayer games, and they are permanent, so if you see your opponent has one of them you can know that they have earned it by finishing in the Top 50 in competitive play. The Cribbage Pro Gold awards are as follows (not cumulative):

  • 4 Gold for Top 50
  • 7 Gold for Top 25
  • 10 Gold for Top 10
  • 20 Gold for 3rd Place
  • 30 Gold for 2nd Place
  • 50 Gold for 1st Place
Remember to complete your 10 matches per weekSee the FAQ for details. Play early, and play often!

Here was the final Top 50 for Autumn 2021:


1ThreeG264MrHand
2iPeg27glaicer29
3Dhrun28jkruger
4card5529ernie313
5flashmatt30gstorm77
6kcguy31Paxter
7Saskie32RatKing
8lbolt5833VGKnights
9WhyADuck34dkatz1877
10XLNC2135Linkup
11UberPooch360wl
12Domerzag37Br1Guy
13nob4one38cribberoo
14jwick6039jwr13
15hillchem40dph
16xdb113541Yompopo
17rattle1542DdsG
18nvsru91143shudbgolf
19chiyo144stephzzz
20beth022345gmax
21Cmoney42146airmark7
22jjonell47Trucha1
23mountains48Rhuby
24better6949Birddoggy
25Swodis50lovJesus


Important Changes This Season

Now to briefly go over some important changes this new season. First, we have changed the rule about how late you can come into the season and start playing. You now must have completed your placement matches 3 weeks prior to the end of the season. This should give enough time to let the ranking settle out the top players without the potential for a few lucky matches to drive rankings sideways at the end of a season.

Second, multiple pieces of the matchmaking system have changed. The biggest change is that we now will enforce a maximum CR difference between players once you get to the Diamond level. That means you shouldn't see someone who is 1000 CR lower/higher than you being matched at that point (and it is fairly unlikely at the lower levels too). The second change is a bit more complex to describe, but put simply it tweaks how the system determines if it thinks the matchup will produce a "good game" and let it through as a positive potential match. I think both of these changes will allow some more movement in the leaderboard in a good way, while also reducing that "risk" of waiting too long and being matched against someone too much lower than you. One of the real life "daily" impacts is that you may wait a little longer for a match, but also that you should not exit out of the matchmaking while waiting like before if you were trying to avoid that lower CR match. It will be fun to see how this all plays out, and note that we may make some tweaks throughout the season if it looks like it any of this isn't achieving the desired outcomes.

One last minor change is a tweak to how wait times are calculated. It is definitely far from perfect, but just remember it is based on what players of your similar level are seeing as wait times on average recently. As such, it will lose accuracy during slower times when fewer players are playing. I am honestly debating if this feature is very helpful, and so please let me know your feedback if you would rather that space on the screen be used for something else or just gone.

Finally, know that we are looking into some potentially bigger changes for the next season. If you have played a few seasons here and have some thoughts on what you would like to see done, please email us at support@FullerSystems.com with your suggestions. Similarly, if you have never played and something is holding you back, let us know. It is primarily what you all suggest that drives what I work on next, so let your voice be heard.

41 comments:

  1. I thought there was a minimum
    Kcguy played about 30 games.

    ReplyDelete
    Replies
    1. There is currently no explicitly set total minimum games required other than minimum per week. The new change to require no new players for the final 3 weeks does effectively introduce a minimum of 30 in that sense.

      Delete
    2. Thanks for the info but would you mind explaining that again. I'm not entirely sure what you meant exactly. Do we still need to play a minimum 10 games per week, Saturday to Saturday?

      Delete
    3. Yes, you still need to play the 10 minimum per week.

      Delete
  2. So what happens in this new season when I'm playing my first match and I'm about to with and my opponent disconnects? It shows me at 0/5 currently?

    ReplyDelete
    Replies
    1. Yes, this is the same as all previous seasons. Due to some potential ways to try to cheat the system in placement matches, a win by disconnect doesn't count towards placement but is still otherwise considered a win and your future CR and position is still adjusted appropriately.

      Delete
    2. What is the downside or upside of disconnecting during a replacement match I don't understand why people keep disconnecting when they're about to lose

      Delete
    3. @loser, great question. There is no "upside" for the person who disconnects. It counts as a full loss plus penalty for them, it still counts as a (not good) placement match for them, and they have to wait 5 minutes before they can play again. Either they somehow still think they are getting away with something (they definitely are not), or they just are being sore losers and likely no amount of penalty will stop them from doing it. Unfortunately, it is not those people that often contact us, so I can't ask them directly what they are thinking when they do that. If they were to hang out and complete the game, they would have a better placement, lose fewer points and not have the 5 minute wait penalty either. Clearly the action isn't a rational one.

      Delete
  3. Best of 1 matches instead of best of 3. We'll have a lot more players, with more matches, and more movement in the rankings. This season's winner was pre-determined weeks ago and look at his comparatively low match count.

    ReplyDelete
  4. Also and most important, thanks for the game and the opportunity to play. Your support/customer service is awesome and you genuinely listen to the players who provide feedback. Thanks again.

    ReplyDelete
  5. First, thank you for developing and hosting this. Now that there has been over a years' worth of tournaments and multiple seasons, you have more than enough participants to establish a minimum number of matches to qualify to be in the Top 50. 30 matches is not near enough as others have noted. I highly recommend a minimum of at least 100 matches which would be a better indicator of skill over luck.

    ReplyDelete
    Replies
    1. Perhaps, we will see. Ideally any minimum would scale with the time the season has been available. We have some ideas, but appreciate the suggestions.

      Delete
    2. I agree about a higher minimum number of matches unless a meaningful confidence factor can be applied to players with a low number of matches. Cribbage matches are like biased coin flips. A coin that is biased to come up heads 53 percent of the time has a fairly good chance of coming up heads 30 out of 40 times, but a much lower chance of coming up heads 300 out of 400 times. More matches played means a higher confidence in skill assessment. Perhaps the season minimum could be at least 130 matches, approximately 10 for each week of the season, regardless of when the player begins the season.

      Delete
    3. @Sawyer, quite possibly. I am debating something like that which enforces a number per week the season has been running. Then if someone wants to come in late, they can, they just need to make up for missed games. Still looking for feedback from more players as well, so thanks for the input!

      Delete
  6. First, thank you for the work in making this system. I love the game of cribbage and the competitive match system has given me an opportunity to play against other players of similar ability. Ultimately, that us what I find the most enjoyable.

    I generally do not like to critique others work, especially when I feel it is already a strong system. However, since you specifically ask for feedback I would humbly suggest a few thoughts.

    First, I would say more should be done for our non-master players. For instance, I would suggest as a player moves up in your current ranking they play only those in there ranking or perhaps +/- 1. It sounds like you are moving in that direction. If so, then those at the lower rankings could get more time to play their cards. A handicap I think would entice many of my friends to play. Then, as they go up the rankings to different levels less time could be given. Also, if they are playing within there ranking other recognitions besides a top 50 overall could be utilized.

    Second, I would advocate to keep things as a series. It is the best method to deal with luck without having to play a huge number of matches. I hypothesize under a 1 game system your ranking would be similar to what you had a few years ago in mult-classic where a loss subtracted from your ranking...albeit for a season. Those that play a huge number of games will be rewarded. I understand that for many playing an average of 40 matches (120 games) or more per week is feasible, but that just isn't my reality and many other players.

    In conclusion, I most likely won't play this season due to time constraints. But, I will again when I am able. Thank you for the time and effort into the product you have produced.

    ReplyDelete
    Replies
    1. Best of 1 is by far the most popular format in the Classic division, it's what the top players play along with most others. Nearly all of my friends who play won't play Competitive because it simply takes too long to finish a match. Sometimes I only have a few minutes at a time for a quick game and can't commit 10 - 15 minutes for a match. I'd play a lot more Competitive matches in a best of 1 and luck balances out. Look at how most of these seasons end up with a known winner weeks before the season ends because you have to play 2x-3x as many games to catch up. For me I'd like to see a larger player base in Matchmaking which helps ensure finding players who are similarly rated and in a lot less time. I'm sitting out this season because it's too much of a grind to commit to best of 3 matches each and every time.

      Delete
    2. I agree one game matches are the most popular, but the best method in that scenario to deal with luck is to simply play more...many would do just that...those with the time. As a result, I believe the season would just result in a war of attrition....cribbage style.

      Delete
    3. It takes about 5 minutes to play 1 game of one match. As it is right now most in the top 50 at the end of the season have played well over 10 matches per week on average. Riggt now a player can compete for the top 50 who has less time to play as many matches and who approaches it as a less is more method and slows down to play. Those averaging more than 50 matches per week to get to 500 or more matches per season are spending around 107 minutes per day playing 50 matches per week or 150 games. If it becomes a 1 game series people would be forced to invest at least that much time, if not more to compete for the top 50. No thanks...

      Delete
    4. Wow.... I never did the math. I get in about 10 to 20 per week and have a tough time getting to ten at times. Getting to top 50 is a goal of mine, but not at that time investment. I like the current format. If we went to a one game series it seems like the winners would be those that devote some serious time...lol.

      Delete
    5. .I have made it into the top 50 before. It is a good goal for me. I played about 120 to 150 games averaging about 10 to 15 per week for three months. I believe that is a sufficient time investment. I understand some on here feel that is too low a number of matches, as refenced by those pointing out last seasons winner who had about the same as me. I believe that person has now won it twice too and has much better overall stats than me for competitive play over around 1500 matches. The series is meant to decrease the luck factor without having to put the time investment in that some are able to do. The three games with 10 match minimum for three monthsd its job.

      Delete
  7. This comment has been removed by the author.

    ReplyDelete
  8. After 6 seasons, experienced players have figured out a game plan to finish at or near the top of the rankings. Large point values are assigned to the winner in early season match wins. The game plan is to play early, play often, hope you get good cards, accrue large point totals, smile at your good fortune......and then sit on your lead. Current leader and solid player Kev8888 has won 22 of 26 matches and sits at a CR level of 4424. Past season winners have finished at a CR level around 4300 or 4350, even lower. Kev8888 got to this level after only 3-4 days into the current season. I think he could win this season simply by playing and winning about 51% of the 10 matches each week, 4424 CR level is a high bar for other players to reach as the season progresses. This strategy is becoming a pattern. ThreeG won last season using the same approach, congrats to him. I am not criticizing these players, I would do the same thing. But, as seasons progress, it is apparent the system does need some tweaks. It is not competitive when a player can reach an unassailable (if played properly) lead so early in the season. I am in favor of keeping 2 of 3 matches or even upping it to 3 of 5 with fewer minimum matches each week. Maybe another competitive match league could be created for those who are unable to commit to large number of matches? Or, have you considered a tournament style league where players are ranked, qualify and then progress to different playoff levels? Thank you for the hard work, much appreciated. Playing others with the same or better skill levels has improved my knowledge and my game. I look forward to future seasons.

    ReplyDelete
    Replies
    1. Hi, great suggestions and definitely options we have/are considering. Just to clarify, there is nothing particularly special about the early part of the season. The reason you see higher changes in CR early on is simply that when you start out the underlying factors are more loose and allowed to fluctuate more as the system develops its confidence factors about your ranking. You can start a new season even after several weeks into it and experience the same result. As the system gains confidence in your ranking, it becomes harder for it to change (up or down).

      Delete
    2. Nobfor1 the key is to get to 4300ish and slow down...this needs to happen anywhere above 70 games. The later the better because you will move down less with a loss. That is what the ones near the top who stay there do at the end of the season. Some players don't care...those that do slow down. It is also why playing a ton of more games means you can get to 4300ish easier...luck is less involved...for the top players. In that skill is definitely involved. You can't stay there by playing a ton because you needed 3 wins to each loss or more if the person you lose to is far enough below your rating. The winner of each season followed the same formula and he top fewish did that last season. So...I do think the system is working. Same players each time pretty much. The cream seams to rise to the top.

      Delete
  9. I'm getting a tad frustrated at the mismatching of some of my more recent games this season. Example.... I managed to make it to 3900+ points then was matched against a player with 3500+ this resulted in me losing the match along with 40 points. In the next game I was matched with someone on 3200 and then subsequently on 3400. Fortunately, I won those two matches, however, having won two out of three matches I was still left with a lower overall score. I don't mind losing matches but after struggling to obtain a high score and then to see it depleted in a mismatched game is very frustrating. I wouldn't mind being matched against someone with a 100 point difference up or down but 800 points is just way too much. That said, many thanks for game and for all the effort you put into keeping us entertained. Apart from the mismatching issue I can't fault it. Keep up the good work.

    ReplyDelete
    Replies
    1. Hi, so the CR limit we introduced this season is at a difference of about 1000. The CR you see is the output of a formula involving several underlying factors. The system runs a complex algorithm to compare these values and determine the expected quality of the match between two players. This means that a difference in CR isn't the primary factor of matchmaking, and doesn't mean that a difference of 400 CR is necessarily a bad (or good) match. This season I expect to see matchmaking across a broader set of CR values (within the limit mentioned), so don't assume that means it is a bad match.

      Delete
    2. Your comments are extremely embarrassing Paludeja in my humble opinion. This is a 3 month season where 'hopefully' the alogarithms even out the luck involved and skill has the final advantage at the end of the season. To play people beneath you as you state in your ' present' position...very embarrassing...I think everyone should be able to challenge all. I was rated higher than you for the past two seasons...I'm currently outscored massively on most games but holding 50%...it is not about the ratings but about enjoyment and the challenge...to think you're somehow elite at your rating is a joke...all should play all...the cream will rise over 3 months.

      Delete
    3. Yeah, that's frustrating. In the last season I was also matched up against players who were 1,000 CR lower than me and you're right, when you lose you get docked a lot more points than if you win. If they're going to allow that much of a CR discrepancy then the reward for winning should be the same for losing.

      But here's a tip, if you don't want to get mismatched with your CR, cancel the matchmaking search for a match after about 10 seconds and then restart it. The longer you wait for a match the greater the chance it will find a player whose CR is way below yours.

      Delete
  10. There should be a minimum number of games for the season instead of a week. Some of us work long hours sometimes and don’t have time to play 10 games a week some weeks but can make up those games other weeks. Say a season lasts 12 weeks make the minimum amount of games 120. That way people who come in late have to play the same amount of games as everyone else.

    ReplyDelete
    Replies
    1. I like the idea of a season minimum of 10 to 15 per week for the total length of the season and enforcing that instead. There are some weeks 10 is difficult to get but others I can get in 40. To qualify for the top 50 for said week you would need the minimum for the total weeks to that point.

      Delete
  11. I have constant connectivity with Fuller/CP.

    ReplyDelete
  12. What are the odds after discarding for the crib that my opponent and I had the exact same hands? And you say this is random shuffle? Not the first time this has happened this time I have a screen shot of it

    ReplyDelete
    Replies
    1. The exact same cards? Should be impossible, and if you saw that then it would mean there was a bug sending the card data to your device. Send us the screenshot for review.

      Delete
    2. Hi, thanks for sending us a screenshot of your question about the game. It took me a few minutes to find the game in question, as I didn't know your time zone or what day this screenshot was taken on. In any case, I found the classic multiplayer game you played that completed at 2021-12-03 21:32:49 UTC with these cards and your final score. I realize how frustrating it feels to lose a game like this, and then see cards where you assume it means the system didn't produce a random result for you. It can make it feel like you were cheated in some way. I totally understand that feeling.

      We track all of the cards used in games in detail, as it is later used in deck audits as we have published about before on the game blog and for quality assurance purposes. So, for your reference, the last deck you had in this game was composed of these cards prior to the deal:
      "8C","7D","8H","8S","JS","6D","7S","8D","6H","KD","QS","AS","3S","KC","QH","10S","3H","KH","3D","4S","7C","3C","7H","9S","JH","JD","5D","6C","9D","QC","QD","AC","2S","9C","4H","5S","10C","AH","2D","4C","6S","2H","5H","KS","10D","AD","4D","5C","JC","10H","9H","2C"

      As cards are dealt from the top to pone and then dealer, and then it alternates between players, the cards dealt to each player were as follows:
      "8C","8H","JS","7S","6H","QS"
      "7D","8S","6D","8D","KD","AS"

      These two hands are certainly "similar" in some ways, and those are ways that we see as important in the game of cribbage and as humans where we see numbers increasing sequentially (like 6,7,8) without skipping any, but these are not "identical hands". If both players know the basics of cribbage, it is not surprising that they chose out of these cards to keep the 6, 7 and two 8 cards. The 4 other cards that were discarded are still part of the hand that was dealt to you, even though you and your opponent chose to discard them. Each player used their skill to select the "best" of the cards available. This act of selection is an important factor as to why you saw these two hands the way you did. The other factor that makes them appear more similar than they were as dealt, is that the system automatically sorted them sequentially when in fact they were dealt out in a different order than as shown there on the final screen - meaning you were not dealt a 6, then a 7, and then two 8s (if sequence were to matter in your "odds" question).

      Delete
    3. The odds of any sequence of 12 specific cards (in any/no particular order) appearing at the top of any deck are the same for any and all unique sets of 12 cards. You could pick any other 12 cards and calculate those odds and get the exact same answer, as the odds of 12 out of 52 cards appearing at the top are always going to be the same. It is us humans that place importance on these particular 12 cards in this example because we understand the rules of cribbage and that these cards "score points", while others may score less points or none at all. Again, to be clear, the odds do not change if you change out the cards in your question. Further to that point, something like this happening has a positive probability. What that means is that it is in fact likely that eventually, given enough decks/shuffles/deals, such a circumstance will in fact happen at some point.

      You seem to be implying that "the odds are too small for this to ever happen and still be the results of a random shuffle". However, that claim doesn't follow logically. First, as mentioned above the odds of any 12 cards appearing at the top of a shuffled deck are the same for any set of 12 cards you can imagine. If it were true that any 12 cards appearing at the top were in some sense "impossible" in a randomly shuffled deck, then we would never get any 12 cards to do so - which is logically nonsense as we always have some set of 12 cards at the top of every deck of cards. Second, the odds of some cards appearing in some sequence in a deck (whatever those odds are), and then that event happening, is not in itself evidence that the deck is or is not randomly shuffled. For instance, the odds of getting a 29 point hand in cribbage is pretty low, yet I'm sure we can agree that this does happen in real life on some frequency (even if rarely). So just because something is "rare", it does not mean that it is therefore always the result of a deck not being randomly shuffled. The conclusion you are trying to make does not follow from the facts of the event you have seen here.

      I think everyone would agree that to know if a deck is being randomly shuffled by a computer system you need to actually analyze the output from that system in a way that can be used to determine if it is producing randomized results. This must be done by looking at a lot more than just one deck/hand, just as if I were to show you some other random set of hands from the system and try to conclude that the deck was randomly shuffled because it produced that result - it is not enough information or evidence to reach that conclusion. There is a lot of actual science on this topic, and we cover some of this in our blog posts on the randomness of the deck and how it is achieved, as well as in our blog posts where we have publicly published some of the audits conducted. There are ways to determine if a system is producing randomly shuffled decks, and it is a lot more complicated than "did this one particular hand or set of hands show up at some point". Obviously a lot more could be said here about the nature of randomness and how it can be shown, but I'll let anyone interested in that topic explore it more on their own.

      Delete
  13. How is it possible to have Dkatz, akatz and bkatz all in the Top 50? Are they all of the same family as in brothers, sisters, fathers or mothers? Or is it one person with 3 ID's?

    ReplyDelete
    Replies
    1. Hi, anyone can change their username to anything they want that is available. I'm guessing this is just players having a little harmless fun.

      Delete
    2. I know one them, they are a friend on the app and recently changed for fun I guess. I know Dkatz as well and the two are unrelated and in different countries actually.

      Delete
    3. As the dkatz, i can assure you the othwr two were doppelgangers. Flatrery i hope.

      Delete

Keep all comments clean, appropriate, and only post if it moves the conversation forward. All others will not be published. All comments are moderated/reviewed before you will see it publicly. If you have a support request or a comment not related to this blog topic, please email us at support@FullerSystems.com instead.