At the cost of sounding the whiny-ass-b**ch that I'm trying not to be after Barca's mournful exit to blue pig dogs, I'm going to have to exhume the dessicated corpse of UEFA's Away Goals system. First the facts, let not Chelsea fans kid themselves that they won 3-2 on aggregate, the game was won 2-2 on Away goals, the Torres goal was for all practical purposes after the game was over, one of those keeper-comes-up-for-last-try counter-goals that are more a final whistle than the "knockout punch" a winning goal might otherwise have been. So now that we've established game was won on away goals, time to do some exhuming. The objective is to incentivize away teams to attack, make the game more exciting, and bottom line : reward teams that played better overall i.e in harder conditions (away). I'd argue what it's really done is make the home team in the 1st leg become more defensive, I'd much rather take a 0-0 at home into the return leg than a 2-1 or maybe even 3-1 victory. All 3 objectives therefore mega fail at very first hurdle, and I venture this is so because of 3 main flawed assumptions. 1) Home is inherently a huge advantage, for all teams 2) Easier to score at home and (therefore) concede away. 3) Goals scored under harder circumstances are the best proxy for "better team" overall.
The graph above shows how great a home advantage is just within each league, I find the results even more stark when I consider cross-league encounters like UCL, and Europa (but given these are flavored by having to play under the rule being analysed, it seemed a biased sample.) - for eg: Valencia have lost 24% of their Home games, and 29% of their away games so I've taken their net home advantage as 5%, Averaging this out for all of La Liga gives 23% while for Serie A it's 21%. My point is not only EPL at 14%, but also the league wise differences in number of teams that are actually better away than home - both showing the inherent assumption of Home being advantageous to not be true universally and more a function of football philosophy/style. Notable names in the right hand graph? The 3 league leaders Real Madrid, Juventus (who haven't lost any games so...), Manchester United - and before some smartass says "hey, that proves the best teams do better away hence Away goals is logical", that's a tautology of "The team that played better away, is the better team..because.... better teams play better away.." we're talking about the rule emanating from an assumption that Home teams are at an advantage. Clearly not, in some places more than others. like with Blue english pig dogs. Oh did I mention, EPL is also the only league (of these 3) where teams 11-20 travel better than 1-10 (something I would have found intuitive for all leagues)
2) Scoring and Conceding. I'm taking this separately and not a duh-sequitor from the "Home teams Win" assumption, since the rules prefer a team that loses 2-3 away and wins 1-0 at home. The graph on the left : Blue is Avg Home Goals / Game minus Avg Away Goals / game, red is similar for conceding though. My reading, especially since now overall goals scored / game bias gets removed, Liga teams score .6 goals less away than home, and concede .6 goals more away than home, compared to EPL at .4 - again saying the popular wisdom that teams score more at home, and concede more away is less a factor in EPL than Liga. Serie A is a really nice median column in almost every graph we see, testament to its simultaneously technical yet pragmatic style, but more on that some other time.
3) Away goals for the "Better Team" - Before quite advertantly (whose red squiggle informs me shockingly that it isn't really a word) opening the Pandora's box of mother of all existentialist footballism questions "How do you define a better team", whether in terms of possession, passes, chances etc - all of which are way beyond the intellectual capabilities of someone writing a blog post from office (does that explain all the extensive excel work by the way.... imagine how many points I've racked up by staring at a complicated work sheet of sizeable data tables all day...). But I started my rant saying Away Goals were presumably to incentivize previously defensive away teams to attack more, or in general make the game more exciting. The fact that it's done just the opposite might just render this assumption superfluous, but in the interest of argument, assuming an exciting game is what they were going for - I hardly think goals presuppose excitement. Ironically enough, if excitement was priority 1, why devise a method to eliminate extra time and penalties, I doubt spectators find the excitement of an attacking away team (which never happens anyway) more appealing than extra time and shootouts. So excitement is out. We're back to effectiveness, result orientation and all those nice words that were invented to describe EPL one day, so I'll straight away throw out any hopeful suggestions of draws being decided by possession stats, number of passes, or derivatives thereof.
Barca clearly outclassed the crap out of Chelsea 83% possession and Xavi completing 66 more passes than the entire Chelsea team, the arguments that they deserved to crash out anyway were of meaningless possession, horrible finishing, and ruthless chance conversion by Chelsea, only the latter two of which I agree with even remotely. But at 2-1 up, and 2-2 on aggregate, agreeing that Away goals are fundamentally flawed, and also agreeing that maybe all draws going to extra-time/penalties are good for spectators but bad/unfair for teams, if we had to agree to decide the game based on the balance of the fixture (that doesn't mean possession due to above mentioned "wasteful possession" accusation, and that it incentivizes teams to randomly pass among back 4, though quite how that makes sense when both teams are trying the same doesn't really strike me...) - I suggest not just "chances created" since Barca are not just horrible finishers in this game, but all the time. I suggest goal extrapolation basis chance creation and conversion ratio statistics.... Barca, in all Liga games so far, have 563 Shots, 258 Shots on Goal, and 93 goals. That's a conversion ratio of 17% goals from shots, and 36% goals from shots on goal. Against Chelsea in the second leg, they had 23 Shots and 6 Shots on goal - using either metric it is 3.8 or 2.2 goals respectively. This gets added to Barca so they're at aggregate (taking lower metric of 2.2) of 4.2. Chelsea have 517 shots, 175 shots on goal, and 48 goals. A conversion ratio of 9% and 27% resp'ly, and given their 7 shots, 3 on target against Barca - have notional goal count of .65 or .82 - I'm taking the higher one meaning their overall aggregate is 3.82. Barca wins. Fvck you english blue pig dogs.
Did I mention Bilbao and Valencia tonight need to hope for an away goals progression haha. I'm going for Bilbao through 3-2 on aggregate, Valencia through 4-4 on aggregate, only because I friggin need something to cheer after these last 2 weeks. Away goals FTW!
It seems to me that the data you cited support the general proposition that 'home is an advantage':
ReplyDelete1) More away games are lost in general.
2) More goals are scored at home and conceded away.
However, I do agree that the present system may be flawed in that the response of many teams, especially EPL teams, to the present system is to, instead of score away, avoid conceding at home, which results in an overly cautious approach.
Possession rate as decider sounds interesting. If this be so, 'possession' must be properly defined and be seen openly and in real time.
Measures should be taken to save football from the single-minded, physical, result-oriented game that has become prevalent with the exception of Barca, La Liga and Spain National Team.
home advantage wasn't really up for debate, only the extent of this advantage and how this extent varies depending on football style (for which we've taken the league-wise differences as a proxy).
ReplyDeleteWe're not big fans of using possession metrics for more than just an idea of how dominant or fluid a team was, since we're not big fans of the way possession is calculated (we've even seen methodologies as ridiculous as using proportion of passes)- currently, shots on goal and related attempts look more tangible as stats.