Field quality in ATP 250s

Not all ATP tournaments are created equal.  That’s reflected in the ranking system–grand slams are worth 2000 points to the winner, while other tour-level events are worth 1000, 500, or 250 points apiece.  The top players are required to play grand slams and 1000-point events, and their rankings can reflect their performance at up to four additional tournaments.

The 250s are the bottom of the ATP barrel.  The level of play is still very high, but it’s rare to find more than one top-5 or a few top-10 players at any given event.  Some 250-level tournaments are able to attract better fields because they offer appearance fees, or because their location and timing provides a convenient warm-up for a grand slam or 1000-level event.

What interests me is the variety in the level of play.  When I first started looking at tennis stats, I aimed to improve on the current ranking system, taking into account specific match opponents, thinking that a player might be over- or underrated because he had faced unusually easy or difficult opponents.  So far, I haven’t found much evidence of that, but it is certainly true that some tournaments–even at the same point level–are easier to win than others.

Below is a breakdown.  These are all the 250-level events from 2009.  ”AvgDirect” is the average ranking of players who gained direct acceptance into the draw.  ”AvgQual” is the mean ranking of qualifiers and lucky losers, while “AvgWC” says the same for wild cards.  These last two fluctuate wildly, in part because most of these events only have three wild cards and four qualifiers.  The final columns are the average and median ranking for the entire field.

Date      Tourney           Field  AvgDirect  AvgQual   AvgWC  Average  Median  
20090105  Chennai           32          84.2    206.5   196.0    110.0   110.5  
20090105  Doha              32          50.8    375.5   701.7    152.4    69.0  
20090105  Brisbane          32          70.3    131.8   369.3    106.0    49.5  
20090112  Sydney            28          28.7    111.6    69.7     47.9    33.0  
20090112  Auckland          28          43.9     93.5   375.3     86.5    52.5  
20090202  Zagreb            32          55.2    254.0   526.7    124.3    60.5  
20090202  Vina del Mar      28          70.1    163.4   100.3     90.0    83.5  
20090202  Johannesburg      32          96.2   1073.5   266.0    234.3   117.5  
20090209  Costa do Sauipe   32          74.1    216.5   276.3    110.8    85.5  
20090209  San Jose          32          85.1    260.7   269.0    135.3    77.5  
20090216  Marseille         32          47.5    246.8   213.5     95.2    55.5  
20090216  Buenos Aires      32          65.4    156.5   443.7    112.3    76.0  
20090223  Delray Beach      32          81.0    250.3   704.0    160.6    79.5  
20090406  Houston           32          80.4    180.8   175.3    108.2    89.5  
20090406  Casablanca        32          73.5    205.0   604.7    139.7    85.0  
20090504  Munich            32          62.7    240.2    84.7     98.0    61.0  
20090504  Estoril           32          63.5    163.0   177.3     86.6    74.5  
20090504  Belgrade          28          70.8    136.0   674.0    147.1    97.5  
20090518  Kitzbuhel         32          69.0    140.1   118.3     89.2    70.5  
20090608  Queens Club       56          82.8    156.0   266.2    105.7    77.5  
20090608  Halle             32          45.4    221.2    56.3     73.9    46.0  
20090615  's-Hertogenbosch  32          62.4    157.8  1004.3    162.7    70.0  
20090615  Eastbourne        32          67.7    196.0   306.0    106.1    78.0  
20090706  Newport           32         108.0    204.4   448.0    155.0   125.0  
20090713  Stuttgart         32          51.9    166.6   120.3     76.2    59.5  
20090713  Bastad            28          59.3    113.0   928.7    160.1    71.0  
20090720  Indianapolis      32          93.3    154.3   566.3    145.3   109.0  
20090727  Umag              32          68.0    179.8   170.3     95.1    85.5  
20090727  Los Angeles       28          67.0    155.5   151.3     88.7    76.0  
20090727  Gstaad            32          63.5    288.4   210.0    112.4    74.5  
20090824  New Haven         48          56.7    149.4   108.0     70.6    61.5  
20090921  Bucharest         32          73.2    164.0   506.7    125.2    84.5  
20090921  Metz              28          57.8    311.8   640.5    153.8    73.5  
20090928  Kuala Lumpur      28          46.2    236.5   222.3     92.3    58.0  
20090928  Bangkok           28          55.5    137.6   286.7     94.9    64.0  
20091019  Moscow            32          65.8    140.3   655.0    130.3    77.0  
20091019  Stockholm         32          61.2    269.4   359.0    121.6    72.0  
20091026  Vienna            32          71.1    151.2   103.7     86.7    81.5  
20091026  St. Petersburg    32          77.4    193.8   661.0    146.7    88.5  
20091026  Lyon              32          67.6    312.3   312.3    121.1    72.5 

Most of the strongest fields are indeed warm-ups for grand slams. That’s certainly the case with Brisbane, Sydney, Halle, and New Haven. Kuala Lumpur and Bangkok also draw competitive fields–for many players, one of those is their first time on a match court since the US Open, right before the bigger-stakes events in China and Japan.

The weakest fields are easy to explain, as well. Chennai is out of the way; many top players are already in the Middle East because of an exhibition the previous week, or in Australia prepping for the Aussie Open, so why play in India? Johannesburg is similarly out of the way. Newport is right after Wimbledon.

Indianapolis is in a unique position–the other tournament that week, in Hamburg, is a 500-level event. It’s also an awkward spot in the schedule. Players can start the US Open series with Indy, then LA, then DC (a 500-level), but with other tourneys available in Europe, many of the top men don’t come to North America until August for the 1000-level Rogers Cup in Toronto.

Should our estimate of a player’s ability be more nuanced, in order to consider the field quality of the tournaments they play? That’s a difficult question. It does seem clear, though, that some 250s are much more competitive than others.

Who are Wild Cards?

Almost every ATP tournament has a few entrants designated as “Wild Cards.”  These are guys who either didn’t enter on time or whose ranking didn’t qualify them for a tournament, but got in anyway.  Most ATP events have 3-5 WC spots, and the grand slams generally have 8.

Anecdotally, it’s clear that WCs fall into a four main categories. Tournament organizers usually use WC spots to generate more interest in their event, so there is usually some element of crowd-pleasing in most of these categories.  Note that they often overlap.

1. Local players who didn’t make the cut.  The US Open has almost turned this into a science–it generally awards WCs to the top-ranked US players who did not qualify for the main draw.  Usually “local” means “home country,” but issues of residency can complicate things.  Also, some countries that host tournaments do not have strong national tennis programs so they turn to neighboring countries; thus, a Kuwaiti gets a WC into Dubai, etc.
2. Former top-ranked players who are aging or returning from injury.  The two players who got the most ATP-level wild cards last year were Taylor Dent and Marcos Baghdatis, both fighting back from injury (and, not coincidentally, fan favorites).  Sebastian Grosjean and Gaston Gaudio both got 5.  These seem to be more likely to go to players who have had success at the specific event.
3. Current top players who didn’t enter the tournament.  Marin Cilic got three WCs last year, all when he was ranked in the top 20.  Sometimes a player returns from injury quicker than expected or needs more practice after a couple of quick exits.  Even Federer took a WC once last year; Murray took one this year.
4. Stud juniors.  In 2008, Grigor Dimitrov won the boys’ singles at both Wimbledon and the US Open.  He was rewarded the following year with wild cards into 6 ATP-level events.  We could put John Isner into the same category based on his college success–he got 5 last year.  Dimitrov is a better example, though, since all 6 of his WCs came outside of his native Bulgaria.  4 of Isner’s 5 came in the US, and the 5th was in the Aussie Open, on the reciprocal wild card granted to the USTA.

Some Stats

I found 225 wild cards in my database of ATP-level matches from 2009.  127 of the 225 lost in the first round, and they had an overall record of 169-222 (43.2%).

Once we start to break them down, we might expect that the local boys are less successful.  And we’d be right.  Here are W-L records for guys playing in and out of their home country:

• Home: 41.2%
• Non-home: 46.8%

But…those “non-home” players represent a very mixed bag.  Cilic went 10-3 as a WC in Basel, Beijing, and Vienna and Fernando Verdasco won New Haven as a WC, but Dimitrov went 2-6 and Baghdatis went 3-8.

Let’s break down the numbers one step further.  We’ll assume that if a player was ranked in the top 40 the week of the tournament, he would have made the cut and falls into my third category above.  Here are the winning percentages for the four ensuing groups:

• Home + top 40: 6-1 (85.7%) — only two players, including Roddick, who won Memphis on a WC
• Home + not top 40: 97-146 (40.0%)
• Not home + top 40: 26-11 (70.3%)
• Not home + not top 40: 40-64 (38.5%)

Taking out the guys who didn’t really need WCs, the numbers start to even out more.

How Good Is That?

Most tennis players would rather win more than 40% of their matches, but as a WC who couldn’t make the cut, is that a respectable showing?

As I discovered by accident in an earlier project, wild cards do better than their rankings would suggest.  Put crudely, if you have two matches where the 100th-ranked player is playing the 30th-ranked player and in one of them, the 100th-ranked player is a wild card, the wild card is the one more likely to win.  ATP rankings represent a lot of information, much of which has predictive value, but tournament organizers often draw on knowledge that is not encapsulated in the ranking, whether that be home-court advantage, dominance more than one year ago, or something else.