Rating is only based on the 32 players above and below you… it looks like for a lot of the new events rating wasn’t impacted for 1st place like your example.
In fact, to go further here, removing players is an improvement to the rating stat because these are essentially all forfeits that get counted as head-to-head wins. No other Elo or Glicko based system is doing this from what I know.
In reply to the bullying comment. Where do you think the changes from WPPR 5.01-current were mainly targeting? It seems like this league was a huge driving force behind a lot of the changes AND was one of the biggest causes of the “Americans benefit more from the system” thing.
Modern, afaik, is (or was) at least partly owned by Steve Epstein, who has a fucking trophy named after him that gets played for every year at the IFPA world champs. I don’t think this is bullying Modern because, if anything, they are most certainly in the “cool gang.” More about “protecting the shield.”
I have no vested interest in what happens to/with Modern. What I find remarkable that any post I make with multiple points, there is no comment from @pinwizj regarding the facts just a snipe.
I am wrong in stating that the rating of players adds to the end tournament value? Which in turn then impacts future tournament values?
I am wrong in saying that previous changes had no retrospective adjustments made
I’m also pretty sure this is wrong as well, Josh can confirm but when the database rebuilds, it rebuilds everything! So since we went back all the way to 2015 every tournament along the way as the ratings/rankings changed for each player, that would be taken in account for each event they participated in.
If your concern is about whether this ripples out to all other tournaments, I think the answer is yes. I believe that every day, the values are recomputed since the beginning of time. If you correct an error in a tournament a year ago, the rating and ranking boosts will change for all tournaments afterwords. If you correct a player name, it impacts when they count and it changes things.
This has far more impact on ranking boost than rating boost, as Tim mentions above.
The rating of players does add to the end tournament value, up to 25 additional WPPR’s if you had 25 players all with a 2000 IFPA Rating.
The rating does impact the value of every other tournament that any of those Super League players would have attended over the course of time.
This is the FIRST TIME we have adjusted previously results that weren’t a formula change. To clarify, we make corrections all the time - duplicate profiles, incorrect results, that do then change the Rankings and Rating from every date forward as a couple of people posted while I was writing this. Anytime we make a formula change it’s always been retroactive. For example if we deem that the Rating metric should be based on the next 20 players above or below you for a given results instead of 32, that change would start from 1980 and be recalculated going forward from there. This is the first time we are essentially taking a tournament that played by the rules put in place for that year, and holding them accountable for a future rules change with respect to the endorsement of that tournament.
The question is WHY?
The answer is that we have received enough evidence to believe that this statement I previously made: “This leaves a situation where AT BEST it’s unclear whether a player intends on wanting to participate and have their scores recorded, and AT WORST allows every score from every player that is playing in the building the opportunity to be recorded, and for that interest in participating to be ‘assumed’.” falls on the latter part of the statement.
Now, many tournaments have included players that haven’t met the 50% threshold of participation, and I believe that rule is a good rule going forward. Going back and reviewing results myself, most of the time an annual tournament consisted of somewhere between 10-20% of players that fell into this category.
Super League however had approximately 75% of their players fall into this category for any given month.
The benefit that the players truly “competing” were receiving was INSANE (see the pre and post-nuke stats of those players). As Tim mentioned, a majority of the WPPR’s earned and the IFPA Rating point increases were against players that essentially forfeited having a chance to compete.
Fast forward to the 50% rule now being put into place, and I found it “interesting” that many of these Modern players were suddenly now meeting that 50% threshold of participation. While not hard evidence by any means, it certainly lends support to the theory that scores were being recorded for as many players as possible in order to qualify them for inclusion into the IFPA Rankings. When the threshold was simply to play 1 game, 1 game was recorded for many players with no additional games recorded. Now that the threshold was 4 games, these players had 4 games recorded.
The collateral damage of all the indirect players ‘WPPR profiting’ off of playing against these Super League players with their inflated Ranking/Rating pales in comparison to the benefit the Super League players were getting by potentially exploiting these hundreds of players month after month.
It’s certainly a judgement call for us to make, and like keefer said, Modern Pinball was home for IFPA HQ. If were out witch hunting we certainly wouldn’t start within our own headquarters. It’s about making the most fair decision I can make with the evidence I’ve been presented (with the knowledge over the years that EVERY decision I make leaves a group of people happy, and a group of people unhappy).
Does that hopefully provide some meaningful response to your comments?
Out of curiosity, do you happen to have easy access to the ratings of these players in these situations. Presumably, the glicko system should have been less sensitive to the inflation. It would be interesting to see how close it was to post nuke ranking
Greg Poverelli 1782.15 (1777.17)
Alberto Santana 1816.27 (1777.62)
Sean Grant 1816.67 (1861.51)
Francesco La Rocca 1748.99 (1690.11)
Lee Hendelman 1775.65 (1737.69)
Basci Dinc 1673.75 (1581.53)
Eric Asher 1647.47 (1568.73)
Frederick Asher 1687.45 (1636.42)
Compared to the much larger retroactive rating nuke of 2016… when some superleague players were “perfect” players (rating >2000) I am unsure why this one in particular is such a tragedy
Just to continue the IFPA thought process on the value of these ‘casual players’.
Most of the results in the IFPA database for say annual events are already having these casual players excluded from the calculation and distribution of WPPR’s, since of those 10-20% of casual players, nearly 100% of them weren’t Rated.
For Super League, not only did you have 75% of the players as these ‘casual not really competing’ level of players, but nearly 100% of them WERE Rated (because they were customers showing up every month).
The biggest thing this nuke did was remove those players from adding on all those Base Value points, and slowing down the distribution curve of WPPR’s from 2nd place on down.
The ratings impact (as you can see from the quick analysis) was minimal at best.
In general, it looks like rating was capturing where they would settle pretty well. F. Asher it implies might have been hit harder than his playing shows. Just based on the number, probably better than 499, but might not travel enough
Here’s my understanding of the rationale behind the retroactive adjustment. Please correct as applicable.
Josh becomes concerned that players reported in Modern’s results may not be intending to participate in Super League or even aware that they are.
To be certain that players intend to participate, Josh requests that a signed participation form be provided for each participant for each instance of the event going forward. This doesn’t work out because of reasons, so they come to an agreement to only report players who show up for finals instead. Showing up for finals is a very good indicator that a player intends to participate in the event.
Because we can’t go back in time to request each participant sign a form for each instance in the past, Josh decides to apply a filter of “playing 50%+ of qualifying games” as an approximation for whether a given player intended to participate in those past events. This happens to be a filter that is applied to all IFPA events from 2017 forward, but is not really a retroactive application of the new rule, rather a coincidence that the chosen approximation has the same filtering function as the new rule.
@pinwizj guessing for things people had to qualify for but are now ‘retroactively’ qualified they can’t still enter? Like women’s championships a few things from that adjustment made different people eligible but guessing they can’t enter since the field is set?
LOL yes . . . we’re not about to tell someone that they are NO LONGER QUALIFIED after they have all their travel plans booked
Every year we make a giant announcement about “Closing the year” on about January 5th-8th. After that point in time the previous year’s standings are LOCKED for qualifying purposes (SCS, IFPA WC, Women’s, etc).
Every year (including this year), we’ve had players that realized they had an error in their results in mid-January. Sometimes it’s a duplicate account, sometimes it’s incorrect TGP of an event, but we’ve had that end up costing that player a qualifying spot in SCS.
I guess this falls under the category of “be careful what you wish for”. ;)[quote=“Artimage, post:100, topic:2383”] @pinwizj Won’t you please think of the children and make all the bad things go away?!?
[/quote]
Last week’s rank: 404, today 1141. I guess I have some work to do
FYI, SPL already has the rule that only paid members can play in Wild Card event, and usually members have 5+ tourneys. Per new minimum attendance rules, guests will no longer show up in standings. Rod said he’d bring up to the board a simple revision of “member with 5+ tourneys” or something like that to play in Wild Card finals.
