Improved Ranking System ???


[quote=“cayle, post:80, topic:3064”]
In a world where “Jimbo’s funfest leftover 420 tailgate party phat pin-bong rips” tournament counts towards a world ranking, and a world where there are not consistent rules applied to events[/quote]

I agree with that as strongly as I can. I’d go as far as to suggest a list of IFPA approved formats for tournaments, with bespoke base WPPR points awarded to them - but that’s a whole different discussion.

Or even only count the Eff% in your top 20 comps used for WPPR ranking.


I think there’s value in exploring that Power Ranking Index, or whatever ‘that is’, as an additional system that we can highlight. I think that subgroup evaluation of the top 250 might help show better who is the “most SKILLED player in the world” . . . even if they aren’t “RANKED” the best in the world.

I manually ran a quick simulation on a player I feel is over-ranked (Germain Mariolle - sorry Germain, I used you because I know you can take it) :slight_smile: . . . along with a player I feel is under-ranked (Tim Hansen). These rankings are of course due to the amount of play/geography/life happens situations that both of these guys have.

What they do have in common however is their historical performance against the same group of players. I couldn’t filter out just the past 3 years, I had to use the lifetime stats from our player-to-player record page.

If I had to GUESS, I would assume that Tim would have a higher winning percentage against the same group of high level opponents that him and Germain have both played.

I was able to filter out 66 players out of the top 250 where both guys had played that opposing player.

Here were the winning records:

Germain (.413 winning percentage based on a 428-608 record)
Tim (.527 winning percentage based on a 746-669 record)

That data point at least supports my theory that “Tim is more skilled than Germain”. Calculating that percentage out for everyone I think could be an additional cool data point to look at. If you were Vegas using these as odds, I would say Tim has a 56% of winning a match against Germain. To me that seems completely reasonable, and far more reasonable that using straight WPPR’s as the data point.


I got $5 on Germain.


Home, Away, or Neutral location? :slight_smile:

I like the potential for this Power Ranking Index. It will be used exclusively for head-to-head matches that we can all bet on.


Yeah, any of those. Knowing the odds is purely for betting right?

As a competitor I’m usually in the Han Solo mindset and I don’t want to hear the odds, but they would be good talking points for commentary.


I think it shows just how close any of these matches would be. In a 10 game set, Tim winning 5.6 out of 10 isn’t an overwhelming stat of dominance.

I’m assuming there will be a large group of players that fall in that 40-60% winning percentage against that top group.

Things get more interesting with this stat when you pull the ELITE players.

Cayle’s winning percentage for example is . . . 84.4%
Elwin’s winning percentage is . . . 93.4%


and Elwin’s winning percentage against Cayle is around 70%, which is insane.


but Peter Watt . . . j/k :slight_smile:

When @Shep has some time to actually build this automatically on the back end, I think it’ll be an interesting stat that will likely support Wayne’s argument at a more accurate measure of skill up at the top.


I really feel like a “do me next!” is in order so I can feel the shame of growing up around pinball royalty.

Actually, that leaves an interesting point. My %age is probably booty due to my local community I gained skills with - does that make me a worse player? Were my good events flukes or just good days? Why can’t @pinwizj win a banner?


That is certainly an interesting wrinkle. If the only top 250 I ever played was KME, my winning percentage against the “top 250” would be 28.7%. That doesn’t make my actual skill level any different, but by this metric I’m terrible.

There’s probably some additional math nerd work to be done here where me only playing 1 out of the other 249 top players makes for a weaker confidence of the metric compared to someone that has played 237 of the other 249 players.

Then you can nerd out even more, and determine the “strength of schedule” based on WHO those 237 players are out of the 249. Did you not play the top 12 ranked players, versus not playing players ranked 239th-250th? Because that would/should/could have an impact on what your winning percentage means, and how confident of a metric that is.


As luck would have it, they are both entered to participate in City Champ 6 in December in San Francisco. We’ll get a pool going at the front desk!


Jimbo’s funfest leftover 420 tailgate party phat pin-bong rips

I heard Hayden Harker is going to run this in Eugene next year; no entry fee, but required minimum of five dabs before 1st plunge. And it’s 100% O.G. Kush TGP #blazeit #WPPRWorld #hisnameispeterwatt #automaticbowlbidforeugene #power5citiesoverrated


How different if this than the probability giving from their ratings. Glicko is trying to solve this using all available data, but of course the way it is used in pinball has flaws. I forget how to do the math (and also think you need the RDs).


All I know in following our Ratings metric is you can inflate you rating by playing against a ton of not high ranked players.

For this winning percentage metric, we’re filtering out the noise of all that activity outside of the top 250 ranked player base.

For example, PETER WATT (and I use that example because I calculated him out in a private email to Wayne), his IFPA Rating metric shows him as the 3rd best in the world. There are 7 people in Australia in the top 250, so I was able to run Peter’s “Power Ranking Index” metric based on his records against just those guys, and it’s 73.0%.

That 73% comes in lower than some comparable players at the top.

KME . . .93.4%
Zach . . .89.8%
Acciari . . . 88.3%
me . . .85.1%
Cayle. . . 84.4%
Raymond . . 84.3%


There is another metric already: Rating. Yes, the default is Ranking, so that’s what most talk about…but Rating should fill in nicely for those who aren’t consistently in high profile events.

How is Rating not satisfying that need?


There is a local player in Colorado who I played against several years back that was amazing, always getting huge scores and he even won several small events. One year he even became Colorado State Champ, but alas he wasn’t able to travel to big events to see where he stacked up. I always wondered how he would do, and by chance that was the inaugural year of the SCS and it was held in CO around the Lyons Spring Classic. I got to see him play KME, and although he played well and put up some good scores Keith defeated him handily, so I guess what I learned is that one does have to compete on the big stage against the best to really understand where one would rank.


It seems too volatile. I regularly find myself jumping around by 100-200 rating positions, without having done anything much out of the ordinary.


We got a new No.1 :wink:



How low am I in the Power 100?


39.87% for you against the other top 250 over the past three years. There’s the POWER 100 … And THE REST. You’re part of THE REST :wink: