I don't know about all of you, but numbers really tickle my fancy… especially new types of numbers! While watching the debut of The Golf Channel's coverage of the PGA Tour, I noticed they had a new number on their leader board. I came to find out this was their Win Zone statistical systems which output how the system thought each player would finish in the tournament based on the information it had at hand as well as two years prior.
Looking back on when I first saw this and looking at it today, I can say with almost 100% certainty that any golf fan can pick as well, if not better than, their "statistical system" does. Before I ridicule it any more, let's look at some numbers:
WinZone Reliability - Round By Round
Outright Win Zone Leaders Leaders Round 1 18% 30% Round 2 25% 41% Round 3 55% 67%
WinZone Reliability - Leader in Top 3 of WinZone
Round 1 - 44% Round 2 - 58% Round 3 - 90%
Now I am no math genius, nor do I claim to be. However, in looking at those numbers I have to wonder how advanced the calculations their computers are doing. Let's get a bit of perspective on this, shall we?
The Sand Trap's Fantasy League
As most of you know we are running a PGA Tour Fantasy league on PGATour.com. From the week by week standings, here is the number of times players had selected the winner from each of the seven tournaments to start the year:
The Sand Trap Reliability Tournament Players Winner Picks ---------------- ------- ------------ Mercedes-Benz 14 9 (64%) Sony Open 21 0 (0%) Bob Hope 21 1 (5%) Buick Inv. 19 15 (79%) FBR Open 21 1 (5%) AT&T Pebble Beach 25 5 (20%) Nissan Open 24 3 (13%) ---------------------------------------- Averages 20.7 4.9 (26.5%)
A thing to note on those numbers… those picks were made before any round was played. I would think that if you were to allow a player to make his picks after the first, second and third rounds, the percentages would increase - as the Win Zone numbers do.
Let's look at that again. On average, The Sand Trap users are picking the winner 26.5% of the time. I guess I am putting my foot in my mouth now as Win Zone shows they get the winner in their top three 44% of the time after the first round. Because we don't get the luxury of seeing the first round before making our picks, I'm going to deduct at least 5% from that number. That leaves us with 39%.
Okay… my foot is now officially in my mouth on this one.
Just a Slow Start?
One could argue that the players in our fantasy pool are just off to a slow start. If we were to throw out our two lowest picking weeks (Sony and Bob Hope) we end up with an average of 36%… still not as good as Win Zone with my little adjustment.
We all know there are way too many variables in picking any sport, but particularly golf. The mental state of a player plays right into how well they will play for a given week. If a player isn't as sharp as he normally is, his scores, more often than not, suffer from it. Because of this simple fact I don't think that it is possible to create a statistical application that would predict any golf scores prior to a golf tournament. During, maybe, but that is just taking information that is right in front of you.
In the end a statistical system is the same as our guesses… it can just guess faster than we can along with more accurately.
The Golf Channel may actually be on to a little something, even if it is a bit hokey. Do you pay attention to these numbers when TGC flashes the leaderboard, or do you ignore them all together? I would love to hear your thoughts!
This article was written by guest author Harry Solomon, an active member of our forum.