Rob Minto

Sport, data, ideas

Month: June 2011

The Wimbledon roof myth

There’s been quite a lot of rain already at Wimbledon this year, and that new roof has been wheeled out a few times. Which is great – for TV, and the 15,000 Centre Court spectators.

However, I keep hearing the people say that the tennis is “sorted out” and the roof will keep things on track so there aren’t delays.

Not really. It all depends when it rains. There are 13 days to Wimbledon, with an exponentially decreasing number of matches to be played. The men and women singles are both a field of 128 – which means you need 127 matches to work out the winner for each event. Centre Court can host two men’s matches a day – three if they started very early (between 9 and 15 total sets), and four women’s matches (8 to 12 sets).

So – here’s the tournament plan, the matches required, and whether the roof keeps the whole thing on schedule.

Weekday Day Round Total matches required Will a roof keep the tournament on schedule?
Mon 1 1st round, m&w 64 No
Tues 2 1st round, m&w 64 No
Weds 3 2nd round, m&w 32 No
Ths 4 2nd round, m&w 32 No
Fri 5 3rd round, m&w 16 No
Sat 6 3rd round, m&w 16 No
Mon 7 4th round, m&w 16 No
Tues 8 women qtr 4 Maybe
Weds 9 men qtr 4 Almost certainly not
Ths 10 women semi 2 Yes
Fri 11 men semi 2 Yes
Sat 12 women final 1 Yes
Sun 13 men final 1 Yes

Answer: the roof is great for the last four days, and maybe the second Tuesday. Anything on days one to seven and it’s all about show. The TV audience is happy; the centre court crowd (which is less than half the gate during the first week) is happy; but not the players or the organisers. Or the people with tickets to court one or ground passes.

The lazy journalism of citing Facebook and Twitter

I’m getting very annoyed with the phrase “such as…” in journalism. It’s becoming a lazy substitute for not having concrete facts, and is used especially to write about websites and  social networks where the writer usually has no idea what they are talking about.

In other words, the phrase “such as Facebook and Twitter” really means “something’s going on online and it must be going on on Facebook and Twitter so just say that as that’s the only new thingy we’ve heard of in the newsroom”.

Some examples:

From the Telegraph: “Teachers believe social networking sites such as Facebook and Twitter are to blame for pupils’ poor grades, a study has concluded.”

From the FT (see, I’m not biased): “Banks are searching out fresh ways to engage with customers on social networking sites such as Facebook and Twitter as they look to capitalise on the growing popularity of these mediums…”

I thought the other day – if I had a pound for every time I saw that written… So I googled it. Over 9 million instances of that exact phrase. And then I got curious. What about the other combinations. LinkedIn and, say, Orkut? Why should Facebook and Twitter be the only show in town?

So here’s the grid:

Source: Google results for exact phrase “such as x and y”. Here’s the data.

I’ve given anything over a million a red background, 100,000 – 1m in yellow, under 100,000 white, and put grey on anything less than 10.

So Facebook and Twitter is the killer combo, although MySpace does well with Facebook. But look at the rest – it’s a joke. It’s not as if they have that magnitude of users less than Twitter – or even Facebook.

But the real joke is that Twitter and Facebook are so radically different. To lump them together that often in a lazy format is ridiculous. The network effect of journalists is not far removed from the network effect of social networks, or software, or other monopolistic services. Once everyone starts to say “such as Facebook and Twitter”, there isn’t much point in bucking the trend.

 

The limits of sports stats: the example of Nadal and the WSJ

This year in tennis is been all Djokovic and that winning streak. The narrative of sports is always about who is “the Man”, so therefore, Rafael Nadal must be a spent force.

The Wall St Journal have, they think, proved it. In their piece Nadal Looks Surprisingly Human in Paris they look at the stats of Nadal’s first four matches this year, and compare to the years he has won before.

Nadal’s stats don’t reflect a full-blown disaster. But compared to his first four victories in the years he won here, Nadal is spending on average a half hour longer on court and breaking opponents’ serves far less often. All this despite not playing a single seeded opponent so far.

What’s wrong with this? First off, players who are seeded CAN’T meet another seed until the third round anyway, so that’s hardly stunning. Let’s look a bit more at the stats they cite.

Serve Game WIN% RETURN GAME WIN% GAMES WIN% SET
WIN%
AVG. MATCH LENGTH
2011 85.5% 39.7% 62.8% 85.7% 2:52
2010 85.2% 50.9% 68.2% 100% 2:18
2008 87.8% 65.2% 76.8% 100% 2:05
2007 86.8% 50.9% 68.9% 100% 2:14
2006 81.8% 43.9% 62.9% 85.7% 3:04
2005 87.7% 44.6% 66.4% 92.3% 2:09

Sources: ATP World Tour, Stats Inc.

His average match length is high, but it was higher in 2006 when he won the title – hardly shocking. The only 2011 stat quoted which is worst in the list is the percentage of return games won – 4 per cent lower than the next lowest. Four per cent, which works out at about 2.5 games on the opponent’s serve that he hasn’t won in 4 rounds – less than a break of serve less per match.

So we have boiled Nadal’s struggle down to about a break less per match from 2006, perhaps two per match from his peak, plus a bit more time on court.

It’s hardly evidence of decline. But stats are like that. They don’t always show what seems evident to watchers and commentators. They don’t show the workrate, the struggle on points, the extra deuces, the attitude. Perhaps those things are there, perhaps we’re seeing what we want to see to fit the narrative. Let’s see what happens from here to the final.

© 2017 Rob Minto

Theme by Anders NorenUp ↑