Introducing the Handicap Rating


October 18, 2016 by bsd987

I’ve been collecting stats all year, and while I still need more data and still need to work more to perfect my formulae, I’m ready to show y’all my Handicap Rating, at least as it applies to the European Tour.

The Handicap Rating is a mix of three statistics, two of which I’ve invented. First is Nine-Dart Average. This one I obviously have not invented. The nine-dart average rates scoring ability. Second is Setup Average. This rates players based on how they are at various mid-leg setup shots. I’m still working on this formula, and it might be some time before it’s 100 percent perfect, but I’m using what I have until I work out something better. Third is Checkout Combination Index. This rates how players are at taking out possible outshots compared to an average player. Outshots are judged both by difficulty and frequency: a missed 161 isn’t going to hurt your average that much because it’s both a low-percentage hit and a low-frequency leg winner.

For now, in the Handicap Rating, all three statistics—Nine-Dart Average, Setup Average, and Checkout Combination Index—are rated as equal, or 167 points. The sum of your rating in each is added together to make your Handicap Rating. A player who is exactly average in all three categories is rated a 501. A player who is above average in all 3 would be somewhere above 501. A player who is below average in all 3 would be somewhere below 501. As I accumulate more data, I plan to adjust the formula as well the component formulae so that the Handicap Rating can become more accurate and predictive.

In order to calculate the 167-point subrating, I weigh each statistic to an average player. For example, if the average player has a 9-dart average of 300, and MvG has an average of 340, I’ll rate MvG as about 14 percent above average. This equates to MvG getting a Nine-Dart Average rating of 190, or 23 points more than the 167 an average player would get. I do the same for the Setup Average and Checkout Combination Index.

But the Handicap Rating isn’t just blind statistics: It adjusts accord to form and importance. More recent events are weighted more heavily than older events, and TV events are weighted two times as heavily as Euro Tour events. A player who put in a good run in Vienna in June but then struggled in October at the Grand Prix[1] will likely be rated lower than a player who might have performed poorly in Vienna but upped his game in Dublin.

The rating is primarily aimed at predicting TV events, but a separate Euro Tour-only rating (which you’ll see shortly) can be used to predict European Tour events.

Now, I’m also calculating a time unweighted Handicap Rating. This by itself would be useless, as it would treat, for example, a performance in May in Gibraltar equally with one in Hildesheim last week. But used in tandem with the weighted Handicap Rating, you can see who has improving form versus who has slipped.

Let’s take a look from the Euro Tour Handicap Rating so you can see what I mean.

Player Weighted Unweighted Form Shift
Justin Pipe 519.16 480.70 38.45
Gerwyn Price 527.39 500.68 26.71
Cristo Reyes 503.53 477.88 25.65
Daryl Gurney 530.79 511.31 19.48
Rowby John Rodriguez 479.94 460.79 19.15
Robbie Green 521.87 503.81 18.05
Alan Norris 521.54 507.39 14.15
Mensur Suljovic 571.33 558.06 13.27
Ian White 510.07 497.33 12.75
Michael Smith 519.71 508.74 10.97

These are the ten players who have the biggest improvement when the Handicap Rating is time weighted as opposed to unweighted. This means these players overall are playing better now than they were a few months ago. In the weighted version, the 4 most recent Euro Tour events are treated as equal, but older events are considered less and less important to the point where ET1 and ET2 are discarded entirely. In the unweighted version, the 8 most recent European Tour events are treated with equal importance.

The first thing you see is Justin Pipe. Pipe’s form the last two years has been fairly awful, and that comes across in his 480 unweighted Handicap Rating: He’s quite a bit below an average player. But when we look at his recent form, which includes a convincing performance this past weekend in Hildesheim, he’s up to a 519. He’s improved over 38 points on his cumulative rating from earlier in the year. Note also that Pipe and Rowby John Rodriguez each have missed a lot of Euro Tour events, and thus neither have as many legs to look at than the others. This may inflate the importance of a single good or a single bad performance.[2]

Mensur Suljovic and Alan Norris—two recent winners on the European Tour—also have a higher weighted Handicap Rating than unweighted.

The only real surprises are the presence of Ian White and Michael Smith. But with both having relatively mediocre seasons to begin with, all this means is that they’ve been performing less poorly now than earlier in the year.

Let’s compare that to players who have been performing worse recently than they were earlier in the year.

Player Weighted Unweighted Form Shift
Christian Kist 424.77 446.66 -21.89
Steve West 503.67 516.62 -12.95
Dave Chisnall 529.17 538.22 -9.05
Darren Webster 524.46 533.40 -8.93
Joe Murnan 487.83 495.75 -7.92
Robert Thornton 467.02 474.12 -7.10
Max Hopp 480.52 486.79 -6.27
James Wade 512.08 517.89 -5.82

The most glaring names are Steve West and Dave Chisnall. Let’s start with the latter because he’s easier to account for. Chisnall reached the final of ET3 and ET4. Weighted, those count for less. His rating either way is very good. It’s less that he’s playing much worse and more that he hit a purple patch on the European Tour in May.

But Steve West is the more interesting case. West was brilliant in Dublin in front of the TV cameras, as well as in Barnsley in the Players Championship events, but his form on the European Tour actually hasn’t been that great. He scrapped out a couple wins in Hildesheim after a wonderful display against his brother, but he’s frequently missed heaps of darts at doubles, costing him victories against Peter Wright and Andy Jenkins and bringing down his statistics overall. The very same clinical finishing that knocked Phil Taylor out of the Grand Prix is missing from his recentEuropean Tour campaign. His rating is still slightly above average, picked up by his great scoring and setup play. But overall, he’s not playing as well on the European Tour as he was earlier in the year.

Anyway, this is just an introduction. I’ll post more about all four of these statistics over the next few months in the lead up to the World Championships.

[1] Because of the unique format of the Grand Prix, I’m currently not using Nine-Dart Average for that event. I’m considering a few alternatives but haven’t settled on it.

[2] I’ve only included people who have played 50 legs on the European Tour in the last 8 events. However, anyone with fewer than around 100 legs probably does not have enough for a completely appropriate rating.


2 thoughts on “Introducing the Handicap Rating

  1. TJ Murphy says:

    Hi Burton fascinating work as usual 🙂 Just a quick question….Do you think eventually that if this new statistic becomes a vital part in darts, will players/fans calculate their handicap rating like mental maths or arithmetic? Similar to what we do with 3 dart averages. I really look forward to your in depth explanation in the coming weeks as I am still a bit unsure of some statistics! Thanks again!


    • bsd987 says:

      It would be tough to do, since it’s compared to an average PDC player and requires analysis of throw-by-throw results. If you had detailed throw-by-throw stats of how you did and paid me to calculate your handicap, I could!


Leave a Reply to bsd987 Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: