Back in January, I read with great interest Dan Weston’s article about favorites in the latter stages of Grand Slam tennis tournaments. Interestingly Dan found that blindly betting on every favorite from the fourth rounds onwards at each of the four Grand Slams between 2010 and 2013 would have made a very nice return of 6.17%.
I locked this away in my “come back to this later” vault, but as things tend to do, it was quickly forgotten. That was until I was watching the recent World UEFA Champions League Cup finals, and noticed that favorite after favorite just kept winning. It was then I remembered Dan’s article and wondered, would his results apply to other sports as well?
The more I thought about it, the more it made sense that favourites would do well in the later stages of tournament and leagues. Not only has the market had an extended period of exposed form to analyse, but with few games available to analyse and bet, the focus of the market is squarely on these available matches.
So is this the case? Do favourites actually do well in finals/playoff matches?
To test this out I first looked at the finals matches of the last 5 seasons from the National Rugby League (NRL), and Australian Football League (AFL). Both leagues promote the top 8 teams from their respective regular season competitions to a four week finals series, and since 2012 both leagues have followed the same finals format.
Interestingly over that period, 71% of favorites won finals matches in both leagues. Betting blindly with level stakes on each favorite would have returned a -14% loss for the NRL, while the AFL returned a surprising 12%. This gives us an overall loss of -2% across both sports. Not a very encouraging start for the theory, especially given that blindly betting the NRL underdogs returned -1%.
But if we take out the first round of the finals, and only look at quarterfinals, qualifying finals, and grand finals something interesting happens. Suddenly the NRL’s loss becomes a 10% profit, while the AFL returns 32%. Only count the qualifying finals and grand final and the NRL returns 24% and the AFL 33%!
But it’s the grand final where the favourites really shine, in the combined 10 grand finals only the Sydney Swans in 2012 won as underdogs. Blindly betting the favourite in both the NRL and AFL grand finals since 2009 would have returned a whopping 93%!
|Performance of Favorites in Finals Matches 2009-13|
|National Rugby League||Australian Football League|
|Round Counted From||Fav Win %||Dog Return %||Fav Return %||l||Fav Win %||Dog Return %||Fav Return %|
To test further I looked at the results of the recent NBA 2020/21 playoffs. The playoffs comprise a best of 7 series of matches, with alternating home court advantage.
This gives us the perfect combination of both results without the advantage of a bye week or home ground, while the best of 7 series reduces the impact of “luck” that we may see in a one-off match.
In just one season of NBA Playoffs we also get a similar number of matches to that seen in the combined NRL and AFL results over 5 seasons (86 NBA matches compared to 90).
There are of course quite reasonable and rational explanations that can be made for these results. Quite simply both finals series are geared towards rewarding highly ranked teams from the regular season through “easier” opposition, home ground advantage, and bye weeks. So what if we looked at a sport where some of these factors are taken out of play, would we see the favorites get better the longer the progress of the final?
From the beginning of the first round of the Playoffs, right through until game 5 of The Finals, 61% of favourites won matches, 10% less than that seen in the NRL and AFL. This was certainly not a promising start, blindly betting on all favourite would have lost you 13%, while doing the same with underdogs would have netted a very respectable 23% profit.
But again, as we progressed through the Playoffs the number of favourites winning increased. Counting from the Conference Semis on, 67% of favourites won their matches returning a profit of 6%. From the Conference Finals onwards this increased to 74% of favourites winning for a profit of 16%. So not as high as what we saw with the NRL and AFL, but a 16% return on investment is certainly nothing to sneeze at.
|Performance of Favourites in NBA Playoff Matches 2014|
|Round Counted From||Fav Win %||Dog Return %||Fav Return %|
What also may have skewed this particular season was San Antonio’s upset against the highly fancied Miami Heat in the 5 Finals matches. Only two favourites won in the five games series, meaning that betting on favourites only in The Finals would have lost 42%. This was very different from the results seen in the NRL and AFL so was worth investigating further, as there certainly could be a “Heat factor” at play.
I ran the test again on Finals matches since 2010 to see if the 2014 Finals were an anomaly, or if they would start trending towards what we saw with the NRL and AFL (although it should be noted that the Heat were involved in four of the five series). Through the five series a total of 64% of favourites won their match for a return of 1%, certainly nothing to write home about there, and a long way from the numbers seen in the other leagues.
|Performance of Favourites in NBA Finals Matches 2009-13|
|Fav Win %||Dog Return %||Fav Return %|
I then wondered if the fact that the NBA did have 7 game series meant that team played differently than when defeat meant elimination for the year. It was in these circumstances that we saw the biggest returns across tennis, NRL and AFL. I wondered if running the test over another sport where more than one matches determined the team to progress would produce similar results, but my most obvious choices the Champions League and NHL both have a draw as a betting option at the end of regulation time so making a comparison would be difficult. I also contemplated using MLB, however with the pitcher being a significant part of the market offered I felt that again this would not be comparing apples with apples.
So instead I went to the NFL, where every game in the Playoffs is sudden death. Like the NRL and AFL, the NFL does have home ground advantage to higher placed teams until the Superbowl. Across 55 Playoff matches over the 5 season the favourite won 60% of the time, the lowest seen across all of the sports examined. Like the NBA, betting the underdog in each match would have made a profit (17%), while betting the favourite would have seen a loss of 14%.
However, like all other sports, the favorites re-bounded from the Divisional round onwards, with 66% of teams winning, although you still would have lost 1% of your bankroll. It’s not until the Superbowl that we see something very strange, in the five seasons examined, only the Green Bay Packers won the Superbowl as favorites. Amazingly you would have made 91% profit blindly betting the underdog in each Superbowl since 2010.
|Performance of Favourites in NFL Playoff Matches 2010-14|
|Round Counted From||Fav Win %||Dog Return %||Fav Return %|
So what can we gain from my very brief look into the performance of favourites in finals series? Well unfortunately not a lot, like all matches it is vitally important to handicap each event individually, and a blind betting strategy is unlikely to provide you with long term results.
In collecting the data though it was interesting to uncover what were sides obviously under-rated by the market continuously causing upsets through a finals campaign. Teams like the 2013 Baltimore Ravens, 2012 Sydney Swans, and 2011 New Zealand Warriors won multiple matches as underdogs suggesting that they were not found in betting by many. Like all sport betting it may be about finding an edge and uncovering these gems which will provide the greatest returns in the long term.