• jared@discsanddata.com
  • Siem Reap, Cambodia
Disc Golf
Comparing Tournament Strength Calculations

Comparing Tournament Strength Calculations

Computing my disc golf power rankings is not a super complicated task, but it does have a few steps. I take a player’s placement in a particular tournament and give it a point value. I then adjust that value based on the tournament tier (major, national tour, ‘A’, ‘A/B’ etc. Excludes everything ‘B’ and below). I then alter the value based on how old the tournament is (a tournament loses 4% of it’s value every week after). And finally (not really chronologically), I alter the value based on the strength of the field at the tournament. Then I add that number to a list of that player’s results from other tournaments. For each player I sum their top 10 results. I do that for like 2200 players, sort them up and voila!1

Anyway, there is always tweaking to be done. Actually going through such a big tweak not that I haven’t posted my power rankings in a couple weeks.2 The tweak I’m considering now is to the how I determine the strength of a tournament.

Previously I used a pretty simple algorithm: I assembled a list of the top 20 players in a division (MPO, men’s open, or FPO, women’s open), based on earnings since the beginning of my data, which we’ll call Day 0.3 I would then, for a particular tournament, count how many of those top players competed. If 15 of more were there, I considered the tournament as a top level tournament in terms of strength, and the players point value would receive no strength penalty. If there were 10-14 of the top money earners there, the tournament lost 10% of it’s value. Then for every top player it lacked below 10, the tournament would take a further 10% hit. So something like…

Top 20 Money Winners ParticipatingTournament Strength Penalty Factor
15-201
10-14.9
9.81
8.729
7.6561
6.59049
5.531441
4.4782969
3.43046721
2.387420489
1.3486784401
0.31381059609

It seemed to work alright. It gave me these lists as the tournaments with the strongest fields (tournaments without a year are 2021)…

Rank for MPOTournament NameMPO Strength PenaltyRank for FPOTournament NameFPO Strength Penalty
1Dynamic Discs Open11Dynamic Discs Open1
1Las Vegas Classic11Las Vegas Classic1
1Waco Charity Open11Waco Charity Open1
1Portland Open11Portland Open1
1Texas States11Texas States1
1OTB Open11Santa Cruz Masters1
1Jonesboro Open17OTB Open.9
8Memorial Open.97Jonesboro Open.9
8The Open at Belton.97The Open at Belton.9
8Santa Cruz Masters.97US Womens Championship.9
8Vintage Open.911Utah Open.81
8Mid-American Open.912Vintage Open.6561
13Utah Open0.72913Memorial Open0.59049
14Goat Hill Park0.5904913Goat Hill Park0.59049
15Shelly Sharpe Memorial0.43046715Victoria Open0.430467
15The 303 Open0.43046715Mid-American Open0.430467
15The Open at Belton 20200.43046717Tennessee States.387420489
18Huk Central.38742048917Bowling Green Open.387420489
18Chain Hawk Open.38742048917St. Patricks Classic.387420489
18Bowling Green Open.38742048917Norman Pro-Am.387420489
17The Open at Belton 2020.387420489

So I have assigned a penalty to tournaments with fewer top players. Here’s what the power rankings look like using this criterium to measure tourney strength…

But of course not all top-20 moneyed players are the same. If I had to play for my life against either the Mandujano sisters or Paige Pierce/Cat Allen, I’m getting full on whooped, and Alexis and Valerie are great young players, but… Cat and Paige are world champs. So what if I include how highly ranked players are in the top 20…

So I made the tweak. I took the list of the top 20 money winners, and instead of considering them equally, I gave them points from 20-1 inversely to their ranking (1st got 20, 2nd got 19, etc.). Then I totalled the points for those who participated in each tournament. Any tournament that scored equal to or over 150, got credit for being a fully stacked tournament. If they scored less, their penalty factor was .9 raised to the power of 150 minus their score ,divided by 10.4

So what did we come out with after that madness? Here’s the new tournament strength table:

MPO RankTournamentMPO St. PenaltyPenalty ChangeFPO RankTournamentFPO St. PenaltyPenalty Change
1Dynamic Discs Open11Dynamic Discs Open1
1Las Vegas Classic11Las Vegas Classic1
1Waco Charity Open11Waco Charity Open1
1Portland Open11Portland Open1
1Santa Cruz Masters1+.11Santa Cruz Masters1
1Texas States11Texas States1
1OTB Open11Jonesboro Open1+.1
1Jonesboro Open11US Womens1+.1
9The Open at Belton0.958732+.0587329The Open at Belton0.890567-.009433
10Vintage Open0.853815-.04618510OTB Open0.836012-.063988
11Mid-American Open0.609452-.29054811Utah Open0.603065-.206935
12Memorial Open0.584301-.31569912Vintage Open0.525871-.130229
13Utah Open0.430467-.29853313Memorial Open0.458558-.131932
14Goat Hill Park0.371432-.21905814Goat Hill Park0.421491-.168999
15Shelly Sharpe Memorial0.294587-.1358815Victoria Open0.291499-.138968
15The 303 Open0.294587-.1358816The Open at Belton 20200.288444-.098976
17Huk Central0.291499-.09283317Mid-American Open0.279469-.150998
18Chain Hawk Open0.246278-.14114218St. Patricks Classic0.265128-.122292
19The Open at Belton 20200.243697-.1867719Norman Pro-Am0.256879-.130541
20Tennessee States0.238615NA20Shelly Sharpe Memorial0.246278NA
I computed the previous strength penalties for Tennessee MPO and Shelly Sharpe FPO, but I overwrote them. And they’re not important enough to recompute.

So, wow. Big changes. Similar stories in MPO and FPO. The top tier tournaments stayed at the top and were joined by a couple tourneys that couldn’t quite hit 15 from the top 20 list, but the players who went were bosses. Then there were a couple of second tier tourneys that didn’t move far from their previous scores. Then after the top 10 in each… massive cliff. Mid-tier tourneys lost a lot of strength value. This creates a situation in which results from the top 10 tournaments will have a massive effect on the power rankings. I’m expecting some big movement. I should wait until tomorrow morning to run this, but I can’t wait. Here are the power rankings after the change in strength algorithm…

The changes look… very mild. Let’s take a look at the new and old MPO power rankings. The new one will be on the left)…

And the new and old FPO rankings, again, the new one will be on the left…

On the MPO side, top four stayed put. Klein dropped a few slots as his wins at Mid-America and 303 lost value. Meanwhile Adam’ Hammes’ win in Santa Cruz and fifths at Jonesboro and Vintage appreciated pushing him up a couple slots. Other than those guys the rest of the top 11 stayed put. In FPO, Ohn dropped a couple places as her wins at Memorial and St. Patricks and the second at Goat Hill took strength hits. Other than her, the top 11 stayed the same.

There are a few more stories down the list and in the score differentials, but I found it interesting how little the rankings changed. I’m excited to see what worlds does to the rankings…

1 Of course, I don’t the calculations by hand. I let the pythons and the pandas, and a few squirrels, handle that for me. If you want to see the code I use, you can check out my relatively sparse github.

2 You can read about that tweak here.

3 Day 0 is currently November 21, 2020. I am continually adding data both forwards and backwards in time.

4 .9**((150-x)/10)

Leave a Reply

Your email address will not be published. Required fields are marked *