Having just flown out to Toronto for a long-term project in Canada, I’m having to get used to a few changes in my environment: Tim Hortons has replaced Starbucks, anything good is considered ‘awesome’ but, most importantly, the 5 hour time difference meant I missed both European Champions League semi finals as I sat through mid-afternoon project workshops. Football (Soccer) just doesn’t register over here; the big things are Baseball and Ice Hockey.


I tried to get in the mood on the flight over by watching the film Moneyball – a good example of analytics being used to gain a competitive advantage. Instead of ‘gut feel’ recommendations from talent scouts, the film tells the story of Oakland Rangers, a baseball team, who despite their disadvantaged financial position and ability to attract the top talent, employed deep analysis of player statistics to recruit an optimal and winning team. Its good to see Brad Pitt putting his faith in the numbers and analytics. Moving away from a ballpark, evidenced based recruitment decisions is an example of Workforce Analytics that can be used by all organisations. For example, at Capgemini, we recently helped a client develop optimal recruitment plans by analysing the future evolution of their workforce. Analytics itself is no guarantee of success (as Damien Comolli recently found out at Liverpool…) It’s important to understand the full organisational change impacts – we do this by combining business analytics consultants with HR experts from our Employee Transformation team.

Having landed, I got to the hotel and couldn’t find highlights from the afternoons football. Instead there were at least 5 different ice hockey games being televised as part of the Stanley Cup – the ‘Champions League’ of ice hockey. The match I watched went down to the wire:

Blackhawks Even Series, Defeat Phoenix 4-3 In OT: Chicago down by 1. On the road. Seconds to go. 6 attackers on the ice. Where have we seen this before? For the second consecutive game, the Blackhawks net a late game-tying goal with just 5.5 remaining on the clock coming off the stick of Brent Seabrook and tipped in by Patrick Sharp sending the game into overtime.

Essentially Chicago got a goal with 5.5 seconds remaining by switching (pulling) their goalie for an extra attacker. But they only just managed it with 5 seconds to spare and given they dominated possession and created many more chances playing with an extra attacker, I was thinking surely they should have done this earlier to give them a longer time to equalise. A quick look on the internet and you realise that hockey and baseball are studied by operational researchers and analytic professionals just as cricket and football are analysed by Duckworth, Lewis and your armchair fan with Cricinfo and Opta Statistics in the UK. One paper has studied the ‘pulling the goalie’ strategy to see if the coach’s gut feel is correct:


“If you’re down by one goal, you’re looking at the minute mark”

The paper shows how the “one minute, when trailing by one” rule does not stack up when analysing the statistics. Using poisson processes for goal scoring and analysing the impact of both scoring and conceded goals when playing with an extra attacker and without a goalie, the model shows the optimal to pull the goalie is with around 2 ½ minutes remaining:

I am reliably informed, and I guess its not too difficult to see, that if you told a coach that your analysis was supported by models and statistics, the baseball coach is far more likely to listen than the hockey coach. Probably again like cricket and football. An Ice hockey player will never listen to your statistics, just as Ronaldo doesn’t take note that his best strategy is to shoot the penalty kick straight down the middle

Although I understand little else about hockey, I can now watch the matches with a little more interest as the game draws to a close and as I pound away on the hotel treadmill trying to pick out a silly little puck on the screen in front of me.