Handicap Upgrading

Trying to find the right solution

Even though we're coming to the end of the season, I've been trying to make some progress over the last few days in terms of handicap calculations. I don't know if this Wednesday League will survive... the future will tell us... but it would be a shame to throw away all the work done, all these hours of research and study for the creation of new spreadsheets to try to find a solution that is suitable, that is logical, so, I'm reproducing it here, it will still be useful.

A lot of people in our league talk about it... a lot of people criticize or complain and somehow... it's normal and I'll try to explain where I'm at today.

This year, the first year of our RWPL with the handicap system has been an experimental year, let's not forget that... and yet it has caused a lot of discussion.

First of all, I would like to remind you that this handicap system was introduced to allow weaker players to have a few chances to win a game from time to time, otherwise it would be difficult to convince them to participate in our league.

So, if we help weak players a little on one side, we also have to "handicap" some very strong players to try to tilt the balance a little.

I was able to talk to some "A" rated players who seemed to complain about this handicap which was really hard: giving two handicap balls to a "C" player or a "D" player is really complicated and difficult for them.

Certainly the "A" players have more difficulty to win especially when it happens on only one game... and some "D" players are sometimes very happy and happy to have succeeded in winning a game thanks to this handicap against a player in principle unbeatable for them. 

It is the principle of this handicap system which was not really developed this year, but some seem to have forgotten it. 

It is therefore necessary that the "A" players accept this handicap system as much as the "D" or "C" players accept to take heavy defeats every week.

It is clear that the first updates of the player rankings raised a lot of discussion, but as written above, it was a test, an experiment and over time we manage to find improvements, that's what I tried to do. 

I've just made some "simulation" spreadsheets with the Net Points of all the players and these spreadsheets show the evolution very well.

In the first update, we made the mistake of not taking into account the number of matches played by each player. As a result, players who played only 2 matches were graded higher than they expected, and this is normal.

  • Andrei, Andy, Bang, Boi, Ian, Stas, M, Mimi, Paul moved up a grade when they had only 2 matches under their belt (3 for others).
  • Mick, Rhys, Udo, they went down a grade when they had the same number of matches.
But, ok... it's in the past...

  An idea for an upcoming 10-team championship

  • The principle and schedule of the updates could be as follows...
There would be 4 updates during the season, with the last update serving as a benchmark for the next season...

Update #1 -> after Round 6 the 6 matches played are taken into account.
Update #2 -> after Round 10, we take into account the 6 matches played...
Update #3 -> after Round 14, we take into account the 6 matches played
Update #4 -> after Round 18, we take into account the 6 matches played

Let's project ourselves now into the future

  First Simulation: Grading #1 after Round 1

In these spreadsheets, several parameters are variable and adjustable:
  1. The frequency of updates: see the sketch above. (Rd6 , Rd 10, Rd 14, Rd 18)
  2. The number of matches taken for the calculation (on these spreadsheets I took 6 matches)
  3. The two maxi & mini values that determine the trend of the player's ranking (up or down). For the moment this value has been assumed to be -2.00/+2.00...

Calculation: We take into account the points (+) and the points (-) which gives the "Net Point" .

First Example: Andre: 6 matches played : Net = [-9 / 1 / 1 / 12 / 5 / -13] = -3    ->  -3/24 frames = -0.13  => He is between [-2.00 and +2.00] so his grade doesn't change

Second example: Barry 6 matches played : Net = [11; 16; 4; -2; 9; 14] = 52  --> 52/24 frames = 2.17  => Barry is out of the [-2.00 and the +2.00], so his grade must change from "C" to "B".

  Second Simulation: Grading #2 after Round 10

The calculation will be slightly different for a player who moved up one grade in the previous round... in fact, instead of counting the last 6 games, only the last 4 he played with his new grade will be taken into account.
  • Barry went from "C" to "B" so we take his last 4 results as Net = +1   ->  [1/16 = 0.06].
  • Wen went from "C" to "B" but only played one game in the next round. So there will be no calculation for Wen.

  Third Simulation: Grading #3 after Round 14

It should not be forgotten that in this series of simulations, some Net-Points of some players are no longer very fair, since some have played at a given time with a rank that did not correspond to them.
These are only simulations to clearly explain the principle and avoid making the same mistakes again.

  Third Simulation: Grading #4 after Round 18

This fourth and final update will no longer be used in the current championship... but can be used as a reference for the next season. 
All scores for this season (Rd 1 to Rd18) are updated !