I’m putting it out there: F.A. Cup 5th round weekend is the worst weekend on the football calendar (international breaks aside of course). In the 3rd and 4th rounds there are plenty of matches on offer, whereas in the later rounds not only is the prize in sight, but you have the bonus of Premier League games being played simultaneously. 5th round weekend, though, serves up just 8 matches, and what’s more, they are spread across 4 days. This means that Sunday is confined to Rochdale hosting Tottenham Hotspur…and that is it. There’s not even a decent looking game elsewhere in Europe. Compare that to the Sunday after, when Chelsea travel to Manchester United before Arsenal and Manchester City battle for the EFL Cup, and you see my point.

So with the F.A. Cup taking centre stage (and Togga taking a back seat), I thought I’d delve into one of the most famous clichés in the game: the magic of cup. How often do “giant killings” occur? Are they less common now than in the past? Which teams fall victim to giant killings the most? And, perhaps most importantly for those only interested in the Premier League, what happens to a club the game after they’ve fallen victim to a giant killing?

The Analysis

What constitutes a “giant killing” is subjective; a team rock bottom of the Premier League defeating a team running away with the title could certainly viewed as one, whilst a league two side humbling a championship team would also likely make the list for many people. However, for ease of data collection, I have restricted a giant killing to only the matches in which a Premier League team has been knocked out by lower league opposition. There are limitations to this criteria (when a promotion-chasing Championship side defeats a relegation-battling Premier League side, is this really a giant killing?), but this is probably the best option given the limited time and data available.


Are Giant Killings Becoming More or Less Frequent?

The graph above shows the number of giant killings per season in the F.A. Cup since football in England began back in 1993. As you can see, there has been no change in the frequency of giant killings over the past 25 years, with an average of 5.6 occurring per season. There are a few spikes (1994, 2000, 2008) and a few troughs (1993, 1998, 2007, 2016), but generally, the pattern doesn’t deviate too much. This is interesting because it is often argued that the magic of the cup is fading, but if we take this to mean the number of upsets that are occurring, then it simply isn’t borne out by the data.


Which Team Fall Victim to Giant Killings the Most?

The graph above indicates how often a team falls victim to a giant killing as a percentage of their time spent in the Premier League. That is, whilst Liverpool have been knocked out by lower league opposition on 9 occasions compared to 6 occasions for West Bromwich Albion, given that Liverpool have been in the Premier League over twice as long as West Brom, then the actual percentage is lower for the Anfield club. Of the current Premier League teams (excluding Huddersfield Town and Brighton & Hove Albion as they are yet to complete a full season), Burnley fall victim to giant killings most (twice in their three seasons) whilst Manchester United fall victim least (once in 25 seasons). Interestingly, Wolverhampton Wanderers have fallen victim on all four occasions that they have been in the Premier League…which doesn’t bode well for their fans hopes of silverware next season! Reading are the only team to have played in the Premier League more than once and never fallen victim to a giant killing.


What happens to Victims of Giant Killings in their Next Game?

The graph above shows the frequency of each match outcome in the next league game after a Premier League club has been eliminated by lower league opposition (left y axis) and the difference in average points per game achieved in the subsequent league game compared to what would be expected based on final points total (right y axis). As you can see, a giant killing actually seems to have a positive effect! Teams only follow up with another loss 25% of the time, with a sizeable increase of 0.16 points compared to a “normal” match (this equates to over 6 points if a team were somehow able to play every league game after a giant killing!)


Final Thoughts

The magic of the cup appears to be alive and well, but whilst the number of giant killings hasn’t changed, the specifics of these matches – not analysed here – may tell a different story. In previous years, it could be argued that upsets occurred because teams were on a more even footing. Rich clubs and poor clubs existed, but nowhere near to the scale of today. Thus, it was possible for the 2nd, 3rd, 4th, and 5th placed Premier League teams all to get knocked out by lower league opposition (yes, this did happen; back in 1994). Top sides still get humbled today (Manchester City and Chelsea did so on the same day in 2015), but because of the new strategy of resting players in the cup, the reasons now are probably less due to “the magic” but rather a consequence of the “giant” fielding a weakened side.

So with all that, is there anything we can predict for this weekend? Well, so far there has been 4 giant killings in this year’s F.A. Cup (Bournemouth and West Ham to Wigan, Stoke to Coventry, and Arsenal to Nottingham Forest). This is in pretty close to exactly what we’d expect based on the past 25 years: 5.6 total giant killings, with 2.5 occurring in round 3, 1.8 occurring in round 4, 1 occurring in round 5, and 0.3 occurring in the quarter finals. Which means there should be one unlucky team this weekend that falls victim to lower league opposition…Leicester? Chelsea? Swansea? Brighton? Tottenham? Or Manchester City? Whoever it is, they probably won’t lose their next league game. Silver linings.


If you enjoyed this article you may also enjoy the book “The Inner Geek of Football” available from Amazon.co.uk for £3.99 (or Amazon.com for $5.24). All of Luke’s Togga articles can also be found here.

Follow Luke Wilkins on Twitter using @the_innergeek


A Couple of Caveats:

  • To some extent, the data reflects how the F.A. Cup draw plays out. In some years it may be that there are substantially more top flight matchups, which means there is less opportunity for giant killings. However, collecting data over such a long period should hopefully see these random fluctuations even out, and make the overall numbers more reliable.
  • Data from the 1993, 1994, and 1995 Premier League seasons – in which 22 teams competed as opposed to 20 – has been accounted for. That is, the frequency of giant killings reflects a 20 team league to keep it comparable with the rest of the data. E.g. in 1994, 12 giant killings occurred, but this is listed as 10.91 in the data (12/22*20).