Why Tom Watson Should Be Adding And Not Subtracting Ryder Cup Captain's Picks

March 20, 2013

Page: 1 | 2 | 3 | Full



Tom Watson announced today that he is reducing his 2014 Ryder Cup captain’s picks from 4 to 3. The numbers show he should be doing the opposite.


Over the last 20 years, the captain’s picks have greatly outperformed the players that get automatic spots on the U.S. Ryder Cup teams.





Watson’s decision seems to be based on fairness. He stated, “I think giving players the opportunity to earn a spot on merit is the right thing to do.” While this is a noble sentiment, the statistics suggests the captain’s intuition and subjective judgment may be the better approach. In a way, the numbers suggest the the U.S. should not want him to be fair, the U.S. should want him to be Tom Watson!


Of course this is not to call Watson unfair. It should not be surprising that an experienced golf professional’s intuition and sense can outperform a likely inaccurate, stagnant and arbitrary system of selection based on prize money. If the proposed criteria were based on more advanced statistics, that would be a more interesting discussion.



Interestingly, even Watson’s prior pick when he was Ryder Cup Captain in 1993 shows the importance of captain’s picks. Watson selected the then 51-year old Senior Tour (as it was then called) player Raymond Floyd to join the team. There was only one captain’s pick back then. Granted, Floyd was already a Hall of Famer at the time (inducted in 1989). Even so, despite his Senior Tour wins in 92 and 93, he had only one PGA Tour win over the prior 6-plus years. Watson’s judgment and intuition proved successful and Floyd played great compiling a record of 3-1-0 and thus part of earning 3 points for the U.S. team, including his singles match, on route to a U.S. victory.


Watson in his announcement today also indicated that Paul Azinger, who introduced the switch from 2 to to 4 captain’s picks in 2008, as being “behind this 100%.” I suspect Azinger is simply being a team player because he must know the statistics show the captain’s picks outperform the players automatically put on the team.


Consider the results in the chart below. Note that the last column on the right introduces a new measure I created called the Return On Investment Per Match (ROIPM). This takes into account the flaw in the typical “Ryder Cup record” format which makes no distinction between singles matches and the team format matches. I will describe ROIPM in further detail later but for now, realize that it essentially gives singles matches the weight of a full point; alternate shot and four-ball matches are only giving half the value to an individual player because the player only makes up half of the pairing for such a match.



This result is also not just a fluke because of the small sample size of the past three Ryder Cups. In fact, if one looks at Ryder Cups going back 20 years to 1993, the performance of captain’s picks is even better.





Like making a fire, managing a baseball team and sex, most guys think they can do it better than the other guy. Unlike those pastimes, the responsibilities of the Ryder Cup captain seem a little less involved; however, the decisions that do have to be made may require more nuance and experience (I’m not sure what this says about my ability to make fires).


There are many photos of the captain looking serious in a rain suit with a walkie-talkie and earpiece. Realistically, there is very little left to do once the players are on the course. There certainly seems to be no technical coaching and there does not seem to be much more in terms of strategy. For Ryder Cups, the players need to win. No amount of “captaining” changes the fact that the players need to drive the ball in the fairway, hit irons onto the green and most importantly, make putts.


But you and I know already know that.


In the spirit of Sabermetrics and the new math geeks in sports, let’s do some analysis ourselves on Ryder Cup results. To start, what are the actual decisions a Ryder Cup captain can make?


The Ryder Cup captain’s job appears to boil down to the following:


  1. Makes the Captains Picks;
  2. Decided who to play in the team matches;
  3. Look serious and concerned; and
  4. Don’t pick rain suits that leak.


In fact, the Ryder Cup captain has very little to do on the final Sunday with the 12 singles matches. Unlike the President’s Cup, the Ryder Cup captain does not get even a moderate amount of input to match his players with the players put forward by the other team. Each captain simply puts his players forward in his selected order and the players match up (one through twelve) with the 12 players put in by the other captain. From a measurement standpoint, this may be a good thing. The Ryder Cup records are more random this way because the best players are not always matched against the other team’s best. This randomness in fact serves out purposes introducing somewhat less bias in the players’ records.


This article will focus on what appears to be the most important job of a captian: making the captain's picks.





We looked at the past 10 Ryder Cups going back to 1993. Since that time, Europe has won seven (7) and only lost three (3). It is not surprising that the European players tend to have better records than the Americans. However, it has not been because of the captain's picks.


First, we have to decide how to measure a player’s performance. As noted above, a player that wins a singles match is likely providing more value from an individual standpoint than a player that wins a four-ball (or alternate shot) match with a partner. This of course assumes a player is 50% responsible for a win or loss in a four-ball (or alternate shot match). This may or may not be true but let’s leave the more complicated analysis as to whether a player is unusually good or bad in paired format matches to a later date. For now, let’s assume that ultimately, it averages out that a player contributes equally to a win or loss in his paired format matches.


Historically, a player’s success in Ryder Cup’s is measured by his record in matches i.e. Phil Mickelson and Keegan Bradley each went 3-1-0 in 2012 and they are remembered for their great play together and enthusiasm. They each won three matches, lost one match and had no ties.



The somewhat quieter Jason Dufner also went 3-1-0. However, Jason Dufner’s 3-1-0 was much more valuable, as an individual statistic, than Mickelson’s and Bradley’s because Dufner won his singles match. Mickelson and Keegan lost their singles matches on Sunday. A player’s typical Ryder Cup record does not reflect this difference. Consider: if Mickelson and Bradley had lost one of their paired format matches and each won their singles match, they each would have the same individual records of 3-1-0. The difference? In that scenario, the USA would have won the 2012 Ryder Cup.


A better way to consider the value of a player is to only assign a value of 0.5 for paired format matches and a full value of 1.0 for a singles match victory. Thus, when a player plays 4 matches in a Ryder Cup, we should look at that as playing for and potentially contributing 2.5 points (1.0 points for the mandatory singles match plus another 1.5 points for the three paired format matched each worth 0.5 points).


Also recognize in this approach, if a player halves a paired format match, then the player is returning only .25 points for purposes of his ROIPM (i.e. 50% of the possible .50 actual points for the individual when he halves such a match).


Looking at it this way, Mickelson and Bradley each contributed 1.5 points to Team USA and Dufner contributed 2.0 points. As mentioned above, a half point is a significant difference considering a team needs only 14.5 points to win the Ryder Cup outright (recall the team defending the title retains the title in the event of a 14-14 tie).





Next, we considered the contribution made by a player by using his actual points. The idea is that a valuable measurement will be to measure how successful a player is each time he plays. The ROIPM is based on a scale of a full (i.e. 1.0) Ryder Cup point (i.e. if we wanted, we could also show it as a percentage).


In the examples above, Mickelson and Bradley each returned a 0.60 ROIPM based on earning 1.5 points out of a possible 2.5 points (each played 1 singles match and three paired format matches). Dufner returned 2.0 points out of 2.5 possible points for a better 0.80 ROIPM.


If you recall the 2012 Ryder Cup, the real problem with the U.S. team was the total of 0.5 points returned by Tiger Woods (of a possible 2.5 points) and 0 points from Steve Sticker (of a possible 2.5 points and zero is zero in any stat). While Jim Furyk was seen as failing for the team down the stretch (and it perhaps unjustifiably still bothers him), overall he was actually able to return an ROIPM of 25% compared to 20% for Tiger and 0% for Stricker.



Of course no one matched Ian Poulter’s perfect 4-0-0 record that returned 2.5 out of a possible 2.5 points for a perfect ROIPM of 1.00.


Here’s the actual points and ROIPM for each of the teams for the 2012 Ryder Cup (the captain's picks are hi-lighted in yellow):


[captain's picks hi-lighted in yellow]
[captain's picks hi-lighted in yellow]


Poulter and Colsaerts combined for a .60 ROIPM (3 actual points out of a possible 5)


[captain's picks hi-lighted in yellow]
[captain's picks hi-lighted in yellow]


Johnson, Furyk, Snedeker and Stricker combined for a 0.35 ROIPM (3 actual points of of a possible 8.5). The ROIPM was obviously impacted by Stricker’s terrible Ryder Cup (although you would certainly want him on the team now).


At first glance, this would not appear to conflict with Watson’s idea of reducing the number of captain’s picks to three. After all, anything would be better than Stricker’s zero points. However, it is a shortsighted approach. First, Stricker would likely have been a captain’s pick even with only 2 picks because he was ranked ahead of both Furyk and Johnson at the time plus he had such a positive past experience with Tiger. Second, longer looks back to the inception of the 4 picks in 2008 and further back 20 years shows the captain’s picks do better than the players automatically put on the team.


[click to enlarge]
[click to enlarge]


The last time the US Team won the Ryder cup in 2008, the 4 captains picks produced 5.5 of the 16.5 points (33.33%) and thus contributed proportionally compared to the number of players (4 picks made up 33% of team). Granted, in that case, the captain’s picks produced essentially the same ROIPM of 0.611 (5.5 actual points of a possible 9) compared to the rest of the team with an ROIPM of 0.605 (11.5 actual points of a possible 19). Azinger won.


What is also notable is that the player with the lowest world ranking each year produced 3.5 actual points out of a possible 6.5 actual points for a ROIPM of 0.54, This is better than both the automatic picks ROIPM since 1993 (of 0.44) and since 2008 (of 0.51). If we assume this lowest ranked player would also likely have been the least likely to be picked that year of the 4 captain's picks, then it is certainly not a good decision to drop to 3 picks (I acknowledge that it is not necessarily the case that the lowest ranked player would be picked last).


As noted previously and reproduced again below, the captain’s picks outperform automatic team members significantly for the United States.



Captain’s picks also tend to excel. Since 1993, 27 U.S. players have produced equal to or greater than a 0.75 ROIPM in a Ryder Cup. Eleven of the 27 players have been captain's picks, ie. 40% of the time. Over that period, captain's picks on the U.S. team have only make up 20.8% of the team. It seems that captains are good at picking hot players that may not necessarily otherwise qualify.


[click to enlarge]
[click to enlarge]

[The European captain's pick's similarly excelled but not quite to the same degree as the U.S. team]


It is also notable that the names on this list are relatively modest names as their world rankings at the time reflect. There does not appear to be an "insiders" or "friends of the captain" approach to the U.S. captain's picks. Captains appear to want to win.


Are small differences in ROIPM significant? Consider that a captain puts his players out there 44 times in the Ryder Cup (12 mandatory singles matches and 32 players for the four-ball and alternate shot matches, i.e. 16 matches x 2). If the captain can improve his players ROIPM by a mere 0.023 each time, it means a full point for the team. In other words, the difference between a team average 0.50 ROIPM and 0.53 ROIPM is significant and can be the difference between winning and losing.





In terms of captains picks, perhaps there should be even more. For example if the U.S. captain had 5 instead of 3 picks, then that could mean an extra half points based on the past 20 years. Let’s say that means putting the 2 extra captain's picks out there in about 3.5 matches each for a total of about 7 matches (i.e. the mandatory singles match and 2 or 3 paired format matches each). That is 4.5 possible points out there for those two players (2 singles matches plus 5 team matches (2.5 points) for 4.5 points). The historical .44 ROIPM of U.S. team automatic picks for two players would return 1.98 points. Historically, U.S. captain’s picks for the two players would return 2.52. That’s over half a point.


If you don’t think that makes a difference, 5 of the past 10 Ryder Cups have been 14.5 to 13.5 matches.


There is another material job of the captain: deciding who plays with whom and how often. One of the common criticisms of a captain is if they play a player too often and they appear tired. We will consider this with the ROIPM measurement in the future. For now, that analysis is beyond the scope of this discussion and does not appear as significant.

There is a general consensus and approval of the selection of Tom Watson and his no nonsense approach. To state the obvious, the United States has struggled at the Ryder Cup culminating with the historic singles collapse at Medina in 2012. However, reducing the number of captain's picks is a poor decision by Watson.


The numbers (and Watson's own experience as captain) show that a captain’s intuition and sense of which players are hot and would make good additions to the team outperforms the automatic approach based on prize money. If Watson is worried about fairness, he always has the option to select the 9th player on the list based on the qualifying criterion anyway. While the next captain can change the number back, this decision unnecessarily restricts Watson’s options for 2014.


By reducing the number of captain’s picks, Watson appears to be abdicating leadership, not providing it.


Watson and the U.S. Ryder Cup team should stop worrying about perceived fairness and start worrying about winning. 


Page: 1 | 2 | 3 | Full

Douglas Han