The Boat Race: Weigh-in on heavier Cambridge crew
Some love it, some hate it, but Jack Houghton just loves the numbers it throws up. Here he assesses what the statistics can tell us, and reckons that Cambridge, the heavier crew, will prevail against the odds...
"With the preparation of the two crews so clandestine in the run-up to the event, there is little for the press to report: at least the weigh-in offers something tangible to write about..."
To my wife it signals the end of winter and brings cheer at the thought of outdoor events to come. To my editor it is "the nonsense on the Thames". To me it is my chance to dust off my boat race spreadsheet and try and make sense of the nonsense.
In doing so this year, I've picked out four factors that are oft discussed by the wheeled-out-once-yearly rowing commentators, and will attempt to tell you whether their prognostications are brilliantly wise or pure bunkum.
"It's all about last year"
Whenever creating models that try and predict which factors have the biggest influence on any outcome, it is always useful to have a control to compare your results to. In the model I have created - which aims to identify the measurable factors that have the biggest impact on the outcome of the Boat Race - the control I used was the previous year's result.
In other words, I wanted to know whether the predictions my model was making performed better or worse than simply picking last year's winner (assuming that a relative consistency in the crews and supporting staff would mean that a successful team one year would have an advantage in a subsequent year), or simply picking last year's loser (assuming that a defeated crew would find more motivation to overturn the previous result).
Well, depending on which time period you choose to measure, the previous year's winner goes on to win around 40% of the time, whereas the previous year's loser goes on to win around 60% of the time. However, the results only show a slight significance.
So when commentators tell you that, "It's all about last year", give more credence to those who are suggesting the crew out for revenge will have the edge.
"It's all about the coin toss"
Numerous approaches show little discernible advantage for either station, Middlesex or Surrey, and so, obviously related, there is no correlation between winning the toss and going on to win the race. So when you hear the teams talk up their river position post-toss, don't be fooled by commentators into thinking it should affect the betting, and look to profit by opposing any market moves that seem to think it will.
"It's all about weight"
This year's Cambridge crew has weighed-in 42kg heavier than Oxford. In Boat Race terms this weight difference is significant - there has only been a greater weight difference between the two crews on 12 occasions in the 140 contests covered by my data - and it means that Cambridge has won what is seen by some as an important psychological battle.
Increasingly, though, expert commentators encourage us to ignore the weigh-in and look behind the bare statistics it offers. Last year, James Cracknell did so with success; and this year Rachel Quarrell, the 1991 Oxford winning cox, writes in the Telegraph, "that the insistence on weight mattering above other factors is largely a press creation, rather than a real fact."
Quarrell could be right. After all, with the preparation of the two crews so clandestine in the run-up to the event, there is little for the press to report: at least the weigh-in offers something tangible to write about. However, in terms of the various factors I have looked at to predict Boat Race results, the weight of crews does seem significant, with the heavier crew winning around 58% of the time across the whole data set, with that figure creeping up to around 63% in recent years.
However, other research suggests the picture may be more complex. Although numerous studies have shown that the increased power of heavier rowers more than compensates for the increased drag they create, analysis conducted by Niels Secher showed that, as race distance increases, weight advantage decreases. For Boat Race participants then, racing over a marathon 7,000m - more than three-times longer than the distances rowed in most studies - does the weight become a hindrance in the latter stages of the race?
Well, it looks that way. Plotting winning distances against crew weight advantages on a scatter diagram seems to show that, broadly speaking, the bigger the weight difference, the better the chance of victory, but also the better the chance of not winning by a great distance. Those who want to play the winning distance markets, or those betting in-play, will want to keep this in mind on Saturday at around 5.50pm.
"It's all about the start"
Finally, if you're getting involved in-play, it's worth noting that 85% of boats to reach Hammersmith Bridge first have gone on to win; although, perhaps interestingly, or perhaps just coincidentally, that figure has reduced to just 50% in the last six renewals.
Back Cambridge to Win @ [3.8]