Greatest predictor of game/tournament winning (Why the Attack is so important).

[blue3d]

What factor/statistic is the greatest predictor of a game winner/tournament winner (Why the Attack is so important).

I love statistics and mathematics (as well as football, of course). I crunched some numbers for the Euro 2004 and World Cup 2002 (95 total games). I was looking for some common thread as to why any single team would win another, move on to the quarter and semi's, and eventually on to the final. A lot of this is obvious, but the more you delve into the details, the more interesting it all becomes.

The 3 best predictors of success were (in no particular order), Total Shots, Total Shots on Goal, and Shots on Target % (Total Shots on Goal to Total Shots (TSOG/TS x100) expressed as a %). This is obvious, as offense = goals.


Looking at any single game, these were by far the best predictors of winning. (The first part of this discussion and the included statistics, only includes winning games, or those that were won with penalties, silver, or golden goals (W.C. 2002), not games that ended with a tie.) In fact, for the Euro 2004, a whopping 90.5% of winning teams (talking about any game) had the higher Shots on Target %. And 87% of all winners had the higher (or equal) number of Shots on Target. 71.4% of the winners had the higher number of Total Shots.

Furthermore, the percentage of winners that had the higher of all 3 of these criteria were 56.5%. The percentage of winning teams that had 2 of the 3 criteria was 87%. And the number of teams that had at least 1 of these criteria was a whopping 95.7%.

To verify these results, I ran the same test with all of the World Cup 2002 games. The percentages were similar, but slightly lower. The percentage of winners that had all 3 criteria was 44%, 2 of 3 criteria 74%, and at least 1 criterion: 92%. The percentage of winning teams that had the best Shots on Target % was 70%. The percentage of winning teams that had the higher Total Shots was 66%. And the percentage of winning teams that had the higher of Total Shots on Goal was 74%.

If you look at these 3 criteria at Fifa's World Cup 2002 website: http://fifaworldcup.yahoo.com/02/en/t/st/to/tgf/index.html , you will see that all of the teams that made it to the quarter finals, semi-finals, and finals, were at the top or near the top of the list for these stats. It is interesting to note that Germany and Brazil were both numbers 1 and 2 for Total Shots and Total Shots on Goal for the tournament. Also, Korea, Spain, Turkery, USA, England, and Paraguay made the top 10 for both of these lists. Now go here to the bracket, and notice which are the top 4 teams of that tournament (if you have forgotten): http://fifaworldcup.yahoo.com/02/en/t/s/g.html

So, this evidence just supports the obvious.

Now, where the rest of this discussion takes us is more interesting.

For the World Cup 2002, there were 8 games of 50 (winning games...there were 64 total games, but we are not discussing the ties) that had a winner with only 1 criterion. There were 4 game winners that had 0 of the criteria. That accounts for 16% and 8%, of the games, respectively. (or 12.5% and 6.25%, respectively, of all the games). Some of these games were critical to the tournament. One was Germany's 1-0 win over the USA (0 criteria). Another was Spain's loss to Korea with penalty kicks (0 criteria). Other games of interest were those of France. They lost to Senegal 0-1, with Senegal only having 1 criterion. They Tied with Uruguay, even though Uruguay only held 1 of the 3 criteria. They lost to Denmark 0-2, even though Denmark had 0 criteria. You come to your own conclusion on what that means. Personally, I believe it shows that France was a better team than the results show (even though I am not a fan of France).

Now let's go back to the Euro 2004 numbers for a minute. This is where it gets even spicier.

The number of winners or winning games (of 31 total games) that had only 1 of the explained criteria were 3 (one game actually had the same stats for both teams). In other words, as I stated before, 95.7% of all winners had at least 1 of those 3 criteria.

The number of winners or winning games that had 0 criteria, were 1 of 31 games.

Well, guess which team just happened to be the statistical anomoly?

Well, for one of the games, the Czech Republic won Germany (in the group round) with only 1 criterion.

However, the other 3 games that fit into this category were won by Greece. The first game against Portugal was one of them. They had 1 criterion, which was a higher Shots on Target %. Portugal was higher on the other 2 criteria. The next one was with Greece against the Czech Republic. It had the higher Shot on Target %, tied in Shots on Goal, and was lower on Total Shots.

Finally, the ONLY GAME OF THE TOURNAMENT where a team won, having earned 0 of the criteria, was the FINAL, Greece against Portugal. And this was done with only 4 Total Shots, 1 Shot on Target, and a Shots on Target % of 25%. The 1 and only shot of the 90 minutes is the one to go in the net.

This was the anomoly.

So what does this all mean? Well, first off, I believe the results of all of these games shows that offense is the greatest predictor of game winners. Specifically, the teams that have the highest number of Total Shots, Shots on Goal, and Shots on Target %, will be the ones to almost always win the game, no matter the teams. Second, teams should still be focusing on these skills, because that is what wins it for them. Third, to change to a defensive tactic, just because it was perceived that this is how Greece won, is a fools game. Teams like France, Holland, and Portugal should continue to focus on the Attack, because that is what is proven to win games......especially World Cups. These statistics were based on EVERY game of the World Cup 2002 and Euro 2004. Fourth, Greece was the anomoly, not the rule. Did they win? Yes. Should they have won based on their results? No. They were, as the record shows, the 1 in 31. Are there other things to consider? Of course, but I decided to only cover the topic of these statistics. Is the sport dead, as some of you would cry out in despair? Of course not. These numbers prove that as well. Has the improvements in individual athleticism and team skills for the game progessed so far as to make it too difficult for the truly better team to always end up with the winning score? Perhaps, but doubtful. Just remember the last World Cup winner as you answer that question.

[Gordon EF]

blue3d, an impressive study, no doubt but I can't really agree with your point.

Your main point seems to be that winning teams usually have the best attacking stats, like shots on target etc, therefore it is attacking which wins games so teams should focus on their attack.

I would think a team will get more shots on target BECAUSE they are the better team on the day and so will be in more attacking positions etc. And because, overall, they are the best team, they will win the game.

You point that attacking wins games can seem a bit obvious but it's pretty flawed. I'm not sure exactly of what you mean but having a good attack alone will not win you the game, a good defence will limit the shots of the opposing team, so a lack of shots on target can show a defence to be very good, rather than an attack to be great.

Teams should focus on their attack? Again, what your saying is a little vague but focusing, i.e putting more emphisis on attack as a result of these stats would be incredibly dangerous and a great manager would recognise that a good defence, midfield and attack would help to win the attack stats.

Of course the point of the game is pretty simple, to score more goals than your opponents, thereby, attack better or more efficiently, and this may seem to encourage attacking but defence is equally, if not more important than attack in football now.

Years ago, teams would play ridiculously attacking formations, Brazil used to play with 5 forwards. But would you fancy anybody's chances against a good defencive unit like Greece or even Latvia or Lithuania?

Winning games and tournaments is about getting the correct balance between attack and defence for your side and if you get that right, it's likely you will get more shots etc.

Your figures also fail to recognise the importance of tactics and formations. the Greeks had a clear tactic of inviting the opponents onto them and hitting them hard on the breaks and, especially from set pieces and corners 9where they are strong). This will result in stats which show them to gave less shots etc but still control the game and win.

Cheers, Gordon

[blue3d]

Well, it has been a few days since I did my analysis of those games. I guess I should have just left the stats, and not given any commentary, especially as how it was so late.

I also went back and quickly looked at how the defensive stats would correlate to the outcome. I only had time to look at tackles. Actually, there was a negative correlation, meaning that more often, the team with more tackles in the game, usually lost. If you want that info, I can give it to you later.

Actually, soccer/football is a pretty new sport to me. I don't profess to be a guru. I am just learning about it. When I was younger, I totally disliked it. I believe it was just a lack of knowledge of the sport. Unfortunately, most of the United States is like that. I read a poll in USA today, done by ESPN, that showed that almost 74.1% of all Americans have absolutely no interest in soccer; 10.9% a little, 10.1% somewhat, and 4.9% very. I guess that would be a good topic for another post.

So, back to the topic, I was actually looking for some sort of way to see what determines a winner. Yes, there are many, many people with their own opinions. They are just that, opinions. And yes, I know that there is not any "one formula" to a great soccer team.

However, in an attempt to better understand what are the characteristics of a winning team, I just started analyzing the data. I looked for books online on the subject, but to no avail. All that was available were books that showed the basic stats of the game, but no long term studies. I guess I should post and ask if anybody knows where to find that.

Anyway, the big correlation was not in defense stats, but in offensive stats. Of course a team will have to be defensive, but the winning correlated with the offensive stats. For the Euro, it was extremely showing. In win/loss games, 90.5% of the winners had the higher shot% (shot on goal/total shots x100). Shots on target and Total shots were high as well, both being 87% and 74%, respectively. World Cup stats held close to this, but lower.

The conclusion I came to, was that Shots and Shots on Target were extremely important. Defense is obviously important, but not nearly to the same extent.

I also believe that soccer needs to be changed. This is for a few reasons. Yesterday, I looked at what I thought were the best teams, and checked to see how often they won. I know, this is difficult to quantate, but I did my best. I characterized the better team, by choosing the team that had 2 of the 3 criteria. I chose this, because 87% of the WINNING teams held 2 of the 3 items. Only one item could be considered a fluke. Based on this, I found that the better team only won 59% of the time!!! The rest of the time it was a tie, or a loss (I don't remember the loss stats, but it was about 6 % of the time). When the better team wins only 60% of the time, I believe something is wrong. How many times has the better team just not gotten the goal to go in (hit the bar, miss wide, etc.), and the weaker team wins on a fluke, or the game ends in a tie? It happens too often!!!

A good example of this, in my opinion, was Greece winning the Euro. They only won 3 of the 5 teams they played (60%) and only 66% of all of their tournament games (4/6). How about the other 11 teams they never played? These tournaments are a crap shoot. A lot depends on how the groups are made up, or rather, who is in your group, and who you play down the line. It is not a true way to determine the best team. It is great entertainment, however. I don't even want to discuss that game more. Yes, Portugal lost, and twice at that. I could have given a different example (I gave 3 or 4 in the earlier post), but that was the most recent.

One cannot expect every time that the last team standing get there undefeated, like Brazil did in the last World Cup. However, for Greece to win, there is something wrong with the game. The final game was the ONLY GAME OF THE EURO WHERE THE WINNING TEAM DID NOT WIN ANY OF THE CRITERIA. Statistically, based on the last World Cup and Euro, it happens only 5.3% of the time (5/95 games). Is it the fault of the Greeks? Of course not. They played by the rules, and won. However, their win of the tourney was a statistical anomoly.

So, repeating, if only 59% of the time, the better team wins, something should be changed. There are too many ties. The first thing I would change would be to reduce the number of players on the pitch. With so many people on the field, it gums it all up. Reducing the number of players would better show who is actually the better team. To exagerate, imagine if there were 20 players per team on the field. What about only 6? Beyond that, I have no idea. Maybe a rule change of some sort would help, but I have no idea what would work.

What does it all mean? I don't know. I'm posting to ask you all that. What do you think?

[Gordon EF]

Wow! So many stats.

Please, if I sound condecsending, I don't mean to, but I guessed you were a newbie to the sport. I suppose a lot of American sports can be analysed in terms of stats and figures but, believe me, football cannot be viewed like this.

I've watched hundreds of games and played a lot of football so I'll try and give my views on your stats.

Re: Teams which (win or make?) more tackles usually lose the game. I would expect that. A team on the back foot will always have to make more tackles. As a defender, the games which my team loses are always so much more busy for me than winning games, so much so that I find games which we win can be boring.

RE: Learning about the sport. Trust me, the only way you will truly understand football is to watch games, in the flesh. Watching games on TV will be very helpfull but loking at stats can be very misleading in football.

RE:Stats which show a winner. I'm not familiar with American sports but I presume you can do similar studies with them and find an important correlation. In football, a poor team can beat a great team, there are so many variables that such a study canno always yield meaningfull results.

Your comment on the BEST teams is entirely corect. The fact is, football is mostly played in leagues in which there cn be anything from 30 to 46 games per season. In this scenario, the best team will almost certainly raise to the top. However, in Cup or tournament play, the winner is the team with best form or even luck. Many factors can play a part.

You say that it is the fault of football that Greece won Euro2004. I can totally understand where you are coming from but Cup and tournament play has become so much a culture of international football that teams know what they have to do when entering. Greece knew the way to beat Portugal and Portugal should have known how to beat Greece, both teams knew the conditions to the game. Greece, played to their strenghts (which is defensive play) and won the game and the tournament.

Very often, in leagues, you will find that a more offensive team will win the league. this is because defensive teams are more prone to draws than attack minded teams. In the system of 3 points for a win, draws are a poor result for title challengers.

Changing football, again, I understand where you are coming from and I have seen others say the same thing. ( I could write an essay on this subject) Tactics, formations and positional knowledge etc are as much a part of the game as technical ability, skill etc.

What I will say is, Cups are not ment to find the best team, they are there for the team that wants it most, the team that has the right attiutude for all it's games. How many World Cups has the best team won? Not enough, usually you will find that the team that plays the best football will not win the tournament.

Cheers, Gordon!

[former baller]

Blue3D, Welcome to football. If you run analysis on ball possession I think your world of statistics will bear fruit. Time of possession, and Time of possession to creating a credible chance at goal etc. This will I believe bring your tackles and shots into some perspective!

Ball Possession
Where and how often Turnovers/Giving the ball away occur
A teams ability as a whole unit to keep its defensive shape on both sides of the midfield stripe.

The soccer analysts also monitor statistics on passing success and errors of each player to weight their individual effectiveness. All these are the other statistics that can bring more insight into shots and tackles of winners and losers

[Gordon EF]

Blue3D, Welcome to football. If you run analysis on ball possession I think your world of statistics will bear fruit. Time of possession, and Time of possession to creating a credible chance at goal etc. This will I believe bring your tackles and shots into some perspective!

Ball Possession
Where and how often Turnovers/Giving the ball away occur
A teams ability as a whole unit to keep its defensive shape on both sides of the midfield stripe.

The soccer analysts also monitor statistics on passing success and errors of each player to weight their individual effectiveness. All these are the other statistics that can bring more insight into shots and tackles of winners and losers

former baller, what do you think these type of stats can actually tell you about a game of football?

You get different types of games. What about counter attacking teams? Should they're stats not contradict those expcected?

[blue3d]

Blue3D, Welcome to football. If you run analysis on ball possession I think your world of statistics will bear fruit. Time of possession, and Time of possession to creating a credible chance at goal etc. This will I believe bring your tackles and shots into some perspective!


First off, thanks! I love soccer. It has become my favorite sport.

I'm not trying to make you feel bad or anything, but your prediction didn't hold, at least not for the Euro 2004 (and I bet it would not hold for the W.C. 2002 as well). I quickly just ran the numbers for win/loss games. Only 52.3% of the time did the winning team also have the greater possesion. (not including games that were decided on penalty kicks, but were winners, straight-out). If you include those 2 games from the Euro decided by penalties, then it drops to about 48%. That is inconclusive, and far less predictive (actually, not predictive at all) compared to the stats for shot%, shots on goal, or total shots.
In fact, for 3 or 4 of the blow-outs (5-0 Sweden Bulgaria, 3-0 England Switzerland, etc.) the losing team actually had the higher possesion time.

[Gordon EF]

Football is played on grass, not on paper!

A cliche, but one which rings true more often than not.

p.s. Can I ask why you support Porto?

[Gordon EF]

In fact, for 3 or 4 of the blow-outs (5-0 Sweden Bulgaria, 3-0 England Switzerland, etc.) the losing team actually had the higher possesion time.

That's becuase it is a clear tactic of teams like Sweden and England to defend. In essence, give possesion to the opponents and use it against them. But Sweden and England are still far superior to those sides so they should win.

In tournaments and maybe in European competitions, a lot of possesion is in defence, when no presure is on the defence, maybe why possesion is not a great indicator?

[blue3d]

p.s. Can I ask why you support Porto?


Why is any team a favorite of sombody? I bet for most people, it is geography, and then based on a single player.

To be honest, I have only been following soccer for the past 6 years, and only really intensly for the past 4 years.
I lived in Portugal for two years, and eventually married a Portuguese gal. That alone, was not enough for me to be interested in soccer. I watched only 1 or 2 games of WC 94, about 3-4 from WC 98. Then in 99, my wife offered me a Fifa 99 soccer game for the PC. Well, I decided to learn all of the rules, etc., and was hooked. I gave up the PC game very quickly. Funny as how something so stupid got me so interested in the game.

After that, I watched more and more games. By the time Euro 98 came along, I was so hooked, that I bought a short-wave radio, just so I could hear the games at work. (I was so mad at how the Portuguese and Italians were screwed by the referees on their games with the French, but that's another topic). Then, I was really only interested in Portugal's and USA's games.

Now, for the WC 2002, I stayed up all hours of the morning (here, if I remember right, live games started around 2 or 3 am) watching every game of the tournament. It didn't matter who was playing....I was watching every game and going crazy. That is how I was for the Euro 2004. I watched all of the games that I could (besides work), and listened to the others at breaks I had during the day on my short-wave radio. I saw recaps of all of the games with the Fifa Euro Action Video. I was especially happy that it was taking place in Portugal. They really deserved to have the Euro there. I felt so bad for them when the USA beat them at the W.C., and then were ousted. That is because it was so much more important to them, then to the U.S.

The only thing that made me mad about the Euro 2004, was that it was not broadcast on regular or cable television. It was pay-per-view only! However, they broadcasted one of the U.S.'s World Cup qualifications against Grenada during the same week. Well, you may say that a WC qualification match is important, but this game wasn't (the U.S. won a cumulative of 6-2 for 2 games). Anyway, that was a big dissapointment. Thank goodness I could watch the games on the internet.

Anyway, when I started liking soccer, I didn't have any team I liked. Still today, there are no MLS (U.S.) teams I favor. I guess the only one I could follow would be D.C. United, being so close to home (but I don't follow the team at all).

My wife is a soccer/football fanatic. Being European, it goes with the territory. To give you an idea, after the Portugal/Spain game, her hands had broken blood vessels from clapping her hands together, and her voice was rasping, all from screaming at the tv sceen. She gets more excited during the games, but I am the one that follows the team and the players, and keeps her updated.

I remember my first experience with soccer there (in Portugal), with everyone in the pubs and cafes, watching the games. Nobody was at home, and those at home were all glued to the television...and that was not even a major tournament game! That is a big contrast to Americans with any sport. There are definitely fans here, but by no means comparable to what you would see in Europe (or England).

Well, with my wife being a big Porto fan, it sort of rubbed off. Then I started following them ever since. That has only been about 4 or 5 years now. I am one of the fans that complained that Victor Baia did not make the Portuguese squad (keeper) this year. Also, I thought Scolari was wrong for not picking Jardel for the WC 2002, when he was in his prime.

I am a newbie to the sport, but I am glad I found it. It has brought far more enjoyment than any other sports (watching). And, it has made my wife happy, now that she can discuss the sport, and watch games together. It has been good for both of us. In my opinion, soccer/football is truly the sport of all sports.

[NoSix]

What factor/statistic is the greatest predictor of a game winner/tournament winner (Why the Attack is so important).


Hi, blue3d, welcome to BigSoccer.

I think you are half right. You can't conclude that attack is more important than defense, because you have looked almost exclusively at the relationship between offensive stats and winning. (Tackles are a relatively minor component of total defensive strength, as you discovered.) To demonstrate that offense is more important than defense, you would have to analyze both the relationship between offense and winning and defense and winning, and show that the relationship between offense and winning is stronger.

As an aside, and without going into any detail, I think if you did that analysis you would find that the relationship between defense and winning is slightly stronger than that between offense and winning. The cliche that "defense wins championships" is probably true for soccer too.

You are right that total shots and shooting percentage are significant predictors of scoring goals (offensive strength). A more rigorous demonstration of that fact can be found here (http://www.bigsoccer.com/forum/showthread.php?t=104042) over in the Statistics and Analysis forum, which if you love soccer and statistics, is definitely the place for you! :)

[Gordon EF]

In my opinion, soccer/football is truly the sport of all sports.

100% correct!