Friday 1 November 2013

Scoring after a low round, Flip-Flop or Follow Up?

Kaymer's Dad a Boxer? It took a while for the penny to drop!
As someone who watches more than his fair share of Golf coverage, I like to think I'm relatively fair-minded to the plight of the Golf commentator.  OK it's not Coal mining or trawler fishing but coming up with new things to say must be challenging, you can only say so many times that Martin Kaymer's Dad was a boxer (after some time corrected to his Grandfather!) or Gonzo Fdez-Castano can't putt for toffee.


But there is one commentary platitude I find particularly annoying.  It's the "players hardly ever back up a low score with another low one".  Those who followed the Shriners Open a couple of weeks will remember Andres Romero shooting a 61 in the first round and then managing an 81 in the second round.  Quite a reverse, seeing him go from 2nd to missing the cut, very good proof that I'm actually wrong!  Although in the same tournament, the eventual winner Webb Simpson shot a 64 in the first round and then followed up with a 63.


In some ways the statement is correct.  Following a low score with another one is unlikely, but getting a low score anytime is also unlikely.  A low score is a rare thing, back-to-back low scores are like winning the lottery and then winning the next week.  An exaggeration on the relative chances but the principle applies (perhaps three numbers is a better comparison!).  Say a course is averaging 72, an average player comes out and shoots a 62.  The following day the average player is still an average player, he hasn't drastically improved to the point where he expects to shot 62 every round.  He should expect to shot a 72, that's a ten shot difference, which looks a huge difference but in effect it's just a reversion to mean.  This is what I think makes the "backing up a low round" statement seem plausible.
A low score often means a player's number has come up


Some golf commentators & followers will say those who go low, fall into believing the game is easy, only to be let down by reality the next day.  My own opinion is that a professional player knows the low round was an outlier and will have no greater expectation of the following day's play.  But it's one thing to hypothesize on the psychology; let's have a look at the numbers.


What's a low round?  66's at Summerlin or Torrey Pines are very different things, but for the sake of simplicity I'll say a low round is 64 or less.  I'll look at the PGA Tour since 2001 (up until the end of the 2013 season) and see how the low scorers perform.


First of all here's the frequency of low scores (cummalative. ie 62 & below includes 61,60,59) -
 

Score
Number
Frequency, 1:N
59
3
71,564
60 & below
19
11,299
61 & below
92
2,333
62 & below
367
585
63 & below
1,067
201
64 & below
2,937
73
 Table: taken from all strokeplay rounds on regular PGA Tour events since 2001 (214,693 rounds).


From the 214,693 rounds which were played since 2001, 2937 were 64 & below.  A frequency of one in seventy-three rounds or 1.36%.


So lets have a look at how these players perform or follow-up in the next round.  This becomes a smaller sample as we want to look at follow-up rounds, fourth round rounds don't count as they don't  have a next round (5th in some Bob Hope events).


For the players who've shot 64 or below in rounds 1-3 (1-4 in the Bob Hope's) there were 2,448 from 178,397 rounds.  Now from those 2,448 there were 87 players who followed-up with another 64 & below round.  A frequency of one in twenty-eight or 3.55%.


So in fact the frequency of a low scores go from 1.36% to 3.55%, contradicting the statement of "following-up".  Players look more likely to score lower after a low round.  There's a number of caveats to this number, a player who goes low is more likely to be a better player than the average one so scoring low again is more likely.  Also as mentioned there's also no consideration of easy courses, following-up in a Disney Classic isn't the same as a Players.  But hopefully the 1:28 figure for follow-ups proves my point.


Going through the figures I did find it interesting that there were a number of players who had followed-up (by my criteria) more than once.  From the total of 87 there were 70 players who'd managed it, of which fourteen have done it twice and two three times.


Follow-up Player
Number
Tournaments
Kenny Perry
3
'03,05 Colonials, '01 Buick
Steve Stricker
3
'11 John Deere, '09 Colonial, '09 Bob Hope
Ben Crane
2
'10 Colonial, '03 Bellsouth
Chris DiMarco
2
'08 Childrens, '02 Disney Classic
Chris Riley
2
'10 Wyndham, '06 Texas
David Toms
2
'11 Colonial, '01 New Orleans
Fred Funk
2
'02 Texas, '02 John Deere
Kevin Streelman
2
'10 Bob Hope, '08 Travellers
Martin Laird
2
'10 Childrens, '08 Wyndham
Scott McCarron
2
'08 Wyndham, '03 Las Vegas
Scott Piercy
2
'13 Sony, '13 Phoenix Open
Scott Verplank
2
'08 Children Classic, '03 Las Vegas
Sergio Garcia
2
'09 Wyndham, '07 Tour Champs
Tiger Woods
2
'07 Tour Champs (same tournament)
TommyArmourIII
2
'03 Texas (same tournament)
Zach Johnson
2
'10 Colonial, '08 Texas
  Table: players who've shot 64&below in consecutive rounds on more than two occassions.


What I find quite interesting with this list is the number of short hitters on there.  Crane, DiMarco, Riley, Verplank, Funk and Johnson all fall into this category; whether there's a reason for that I'm not sure.  For some reason I associate low scores with long hitters but this and Jim Furyk's recent 59 could possibly prove otherwise.  But all the same this is an interesting list of players who can follow-up successfully.
Ben Crane isn't one of the more likely low scorers.


Whether this actually means that there are players who can follow-up and those that don't I'm not sure.  Going back to Andres Romero's 20-shot swing, anyone who has followed Romero's career since the heights of the '07 Open at Carnoustie will know he's an erratic player and it wasn't completely unexpected.

But I think I've proved that the "follow-up" statement is a bit of a myth and commentators should probably shelve that one.  I won't hold my breath, I don't think statistical considerations are always at the forefront of the commentator's mind!