Suggestions as to how we/SI could improve match ratings for certain positions.

[Jimbokav1971]

Over the years it seems to have become a trend that GK's and DM's especially, seem to always struggle with match ratings. This in itself isn't that huge a deal just so long as you are aware that the reason that they are getting pants ratings is not because of player performance, but simply because of the way that the game rates players playing in that specific role.

While it seem perfectly the norm now in real life, for players such as Makelele (who seemed to bring it into the lime-light), Matic, Parker, Fernando, Fernandinho, Barry, Coquelin & Carrick, (albeit to a lesser extent) to receive the plaudits for creating a base for others to play, FM seems to be..... well, more than a decade behind in terms rewarding what these type of players bring to the table.

It's all well and good rewarding a Fat Frank, Stevie G or whoever for all the goals and assists they score/create, but for every Fat Frank there is a Makelele and for every Stevie G there is always a Lucas/Henderson/Didi etc.

I know it's hard because we don't know what formulae SI use to create the existing match ratings in the first place, but does anyone have any suggestions as to how these type of roles/players might be better rewarded in terms of match ratings within the game?

I'm not particularly keen on pass % playing a more important role. That used to be key and I didn't like it at all.

I would say that interceptions might be an option, but I don't think that's it either. When I look at match stats, what jumps out at me is the following.

We have Passes and completed passes and key passes.

We have Crosses and completed crosses.

We have Tackles and completed tackles and key tackles.

We have Headers and won headers and key headers.

Then we just have interceptions.

Then we have GK stats (which I will leave for another day).

Then we have shots blocked.

Then we have runs/dribbles, (but not failed dribbles where the dribbler was tackled)

There is no way to track the interceptions that could/should/might have been made but weren't, and there is nothing to differentiate a key interception from a centre-back collecting an aimless hoof that comes his way under little pressure.

Again, not only do we not know how the formulae works in terms of match rating, but we don't know if certain things are included at all. For example, a player with a huge engine that gets up and back all game and puts in a huge amount of work for the team, clocking up a huge amount of distance covered, would be rewarded with all the stats that clubs have in the info that is available to them. Is this of any influence at all in the game? I'm not sure it is.

Does anyone have any suggestions as to how this might be improved?

ps. I wouldn't have thought that a player having a penalty saved should have such a big influence either, (but should be huge for the GK and isn't).

[Karnack]

Something like "Pressure" should be included. A player like Henderson have a big role in making sure he is putting pressure on the player holding the ball forcing them to either make a mistake or make pass that's not going forward or anywhere dangerous at all.

I also think that in a system with short passing and retain possession the DM (depending on role) should also be graded somewhat on how well he stays available for a pass whilst your own team is in possession. Is he making room for himself and giving his more offensive team mates options.

[brawla123]

I just ignore the DM match ratings most of the time. There's games when I see my Anchor Man intercept plenty of passes and give the ball to his surrounding team mates but the guys who get the assists and goals get the rewards. I think it makes it harder for these kind of players to receive awards. Whenever I see the Ballon d'Or shortlist, it's usually attacking players who are considered.

With that being said I think the match ratings are slowly improving but it would be nice to see what actually goes into these ratings.

[Jimmious7]

I guess the main issue is how you rate defensive positioning and influence on space, without necessarily making interceptions. One way I can imagine would be to give positive ratings when a player blocks a passing lane effectively, thus making the opponent decide to pass somewhere else. Of course I don't know how the ME calculates this stuff but a DM should get good ratings if he never allows any passes for example from the AM to the FC of the opposing team.

[GazzP]

Doesn't it sort of reflect real life though? Matic has probably been Chelsea's key player this season but all the media coverage has been about Costa/Fabregas/Hazard because they're the one that produce offensively.

[Karnack]

Doesn't it sort of reflect real life though? Matic has probably been Chelsea's key player this season but all the media coverage has been about Costa/Fabregas/Hazard because they're the one that produce offensively.

But the DMs can still get good ratings after a game IRL, they often do. But media will of course report on goals and more "juicy" bits of football

[sirgiorgio]

Beauty is in the eye of the beholder... I'm a big fan of defensive midfielders and holding midfielders in general, and Didier Deschamps was my hero, but I don't know many people (even among self-proclaimed experts) who can tell a good dm from a bad dm.

Of course, you'd expect some degree of expertise from a game like Football Manager, also because ratings have an influence on many aspects of the game, like how the board and the fans react to new signings (imagine a situation like Chelsea/Matic/Mourinho: how would it be represented in the game?)

It's weird though: in terms of tackles, passes, positioning, and possibly even headers I think there isn't much of a difference between a DM and a DC, yet central defenders usually have higher average ratings.

As for goalkeepers, sometimes it's depressing to see how low their ratings are. However, I think they're part of a bigger problem: we know there are too many shots, ergo too many shots on target, ergo too many saves. If they were rated "more realistically", the worst keeper would have an average of 8.0 at the moment.

[turns88]

Doesn't it sort of reflect real life though? Matic has probably been Chelsea's key player this season but all the media coverage has been about Costa/Fabregas/Hazard because they're the one that produce offensively.

That's not really true. Matic has had a lot of media coverage and plaudits this year (and rightly so).

[Sunstrikuuu]

I know it's hard because we don't know what formulae SI use to create the existing match ratings in the first place, but does anyone have any suggestions as to how these type of roles/players might be better rewarded in terms of match ratings within the game?

Does anyone have any suggestions as to how this might be improved?

ps. I wouldn't have thought that a player having a penalty saved should have such a big influence either, (but should be huge for the GK and isn't).

I'd maybe like to see more granularity by role and duty than by position. An enganche will have fewer attacking runs/game than an AMa because of the differences between the roles; attacking runs ought to have less weight for that role than for attacking midfielder or trequartista. Similarly, a CMd and a CMa don't play the same way and the evaluative factors should reflect that.

This is probably all a bridge too far. I wonder about tackles for DMs; we know the tackle stat in FM doesn't work the way tackling statistics do normally.

[Sunstrikuuu]

I just finished a match against Sunderland. Lucas Romero started and played the whole game as a Central Midfielder (defend); Aaron Ramsey started and finished as a Central Midfielder (attack). Ramsey finished with a 7.1; Romero with a 6.6.

Neither player took a shot.

Ramsey attempted 41 passes and completed 37 (90.2%) with two key passes. Romero attempted 28 and completed 24 (85.7%).

Neither player attempted a cross.

Ramsey was involved in 13 aerial challenges and won 5 (0/2 defensive, 2/6 midfield, 3/4 attack with 1 I can't find). Romero was involved in zero (the only player on the team other than the keeper).

Ramsey won 10/10 tackles with one key. Romero won 12 of 18 tackles.

Ramsey committed no fouls and was fouled one time. Romero committed no fouls and was never fouled.

Ramsey made four interceptions and missed eight. Romero made five interceptions and missed two.

Ramsey ran 9.0 miles (team high). Romero ran 7.6.

Ramsey made 3 attacking runs, was offside zero times and made nine mistakes. Romero made no attacking runs, was never offside and made two mistakes.

Ramsey has a clear advantage in passes made and completed. Ramsey completed more aerial challenges, but at a low percentage. Ramsey did not miss a tackle.

Romero made more tackles at a lower percentage. Romero made more interceptions at a much higher percentage. Romero made fewer mistakes.

I don't know what if anything this means, but data!

[isignedupfornorealreason]

I just finished a match against Sunderland. Lucas Romero started and played the whole game as a Central Midfielder (defend); Aaron Ramsey started and finished as a Central Midfielder (attack). Ramsey finished with a 7.1; Romero with a 6.6.

Neither player took a shot.

Ramsey attempted 41 passes and completed 37 (90.2%) with two key passes. Romero attempted 28 and completed 24 (85.7%).

Neither player attempted a cross.

Ramsey was involved in 13 aerial challenges and won 5 (0/2 defensive, 2/6 midfield, 3/4 attack with 1 I can't find). Romero was involved in zero (the only player on the team other than the keeper).

Ramsey won 10/10 tackles with one key. Romero won 12 of 18 tackles.

Ramsey committed no fouls and was fouled one time. Romero committed no fouls and was never fouled.

Ramsey made four interceptions and missed eight. Romero made five interceptions and missed two.

Ramsey ran 9.0 miles (team high). Romero ran 7.6.

Ramsey made 3 attacking runs, was offside zero times and made nine mistakes. Romero made no attacking runs, was never offside and made two mistakes.

Ramsey has a clear advantage in passes made and completed. Ramsey completed more aerial challenges, but at a low percentage. Ramsey did not miss a tackle.

Romero made more tackles at a lower percentage. Romero made more interceptions at a much higher percentage. Romero made fewer mistakes.

I don't know what if anything this means, but data!

9 mistakes in attack = easily forgiven. 2 mistakes in defence = not.

Least I think that's how it's factored in. My attacking players are the ones who make a lot of mistakes, but it's the defensive ones that stand out and bring the ratings down.

[Sunstrikuuu]

That's a good point. It's tough to tell based on the Analysis screen. It looks, though, like the mistakes are all in pretty comparable locations. Romero's are maybe 5 yards outside the D and 10 yards outside and to the left of the D. Ramsey's are all over, but all in the defensive half, with one maybe ten yards outside where the D connects to the penalty box on the left side, one fifteen-20 yards outside the same position on the right side, and one on the edge of the six-yard box.

[louissiscool]

I think one way for SI to improve match ratings for certain positions would be that different things define a good rating in each position.

[chocolatecoatedballs]

It's a hard one because of the role, the only way that I can see it improving is to have match rating for position. GK, CD, FB, DM, CM, WM, AM, CF. With things they are not expected to do gets a higher rating (eg: a striker tackling, a DM assisting or scoring), what they are supposed to do a mid rating (Treq - dribbling, dm - tackling) high for key (DM intercepts, CF goals). In reverse v low rating if fail to do what is expected (WM dribble, DM intercept), mildly low if they don't achieve what is not expected (CD scoring shot, Dm scoring shot), v v low if they fail in a key moment (DM mistake leads to goal, AM risky pass/throughball not being successful).

My examples may not be perfect, but that's beyond the point, the point being that a player is expected to do certain things in a position, there are key moments for each position and there are things that they are not expected to do but can happen.

[brawla123]

Doesn't it sort of reflect real life though? Matic has probably been Chelsea's key player this season but all the media coverage has been about Costa/Fabregas/Hazard because they're the one that produce offensively.

Matic has received a lot of praise this season actually.

[Guest]

Out of interest has anyone conducted a study on real life average match ratings that are awarded by a selection of media outlets? I ask because I do not recall anyone backing up the commonly held belief that certain positions are hamstrung by the way match ratings are calculated with verifiable real world data.

[Rashidi]

Out of interest has anyone conducted a study on real life average match ratings that are awarded by a selection of media outlets? I ask because I do not recall anyone backing up the commonly held belief that certain positions are hamstrung by the way match ratings are calculated with verifiable real world data.

Stands up and:applause: FINALLY!!!! Match ratings are SUBJECTIVE...yes..there is NO verifiable system ANYWHERE in the WORLD. Squawka ratings are squawka ratings, and when you are watching it on the telly, its the commentators who sit down and hash it out, giving ratings. I know cause thats what I used to do for a living. MOM award for a player is done from the press box and NOT all journalists are asked..better still its affected by the sponsors. The reality is Match Ratings have always been subjective, FM is a simulation and the match ratings here may be a direct result of the team winning, the player scoring, and players generally fulfilling their roles..and its probably more objective than any match ratings system you can find in the world. I have contended since day one that this is a factor of peoples tactics and how well they have nailed the roles to the proverbial cross. If its perfectly nailed in YOU will produce a DM with high ratings. All it takes is one person to prove that it can be done, and the whole argument that ratings are broken is flawed.

[BiggusD]

Both my right-back and my friend's right-back have ratings in the high 7s and many many MOM awards. They are hailed as world-class full-backs. Judging by their attributes, they are nowhere near world-class and I can't spot anything obvious that makes their ratings so high. Their attributes are both very balanced - from 13-15 in most attributes, both defensive and attacking ones. They don't have particularly many assists and no goals.

A ball-winning DM may get very high ratings because he will be involved in so many challenges. A DM meant to provide cover for the midfield going forward and a passing outlet from the attacking third, however, will likely not get high ratings. This particular case could be fixed by identifying how many passes and runs the opposition could NOT make because he was in the way. We could divide Interceptions into Active and Passive interceptions. This would also be a statistic that would help us and the assistant pinpoint the reason that we're not creating anything - a team that is standing off will simply block the space and the opposition players will begin racking up Passive Interceptions.

[Jimbokav1971]

Out of interest has anyone conducted a study on real life average match ratings that are awarded by a selection of media outlets? I ask because I do not recall anyone backing up the commonly held belief that certain positions are hamstrung by the way match ratings are calculated with verifiable real world data.

Stands up and:applause: FINALLY!!!! Match ratings are SUBJECTIVE...yes..there is NO verifiable system ANYWHERE in the WORLD. Squawka ratings are squawka ratings, and when you are watching it on the telly, its the commentators who sit down and hash it out, giving ratings. I know cause thats what I used to do for a living. MOM award for a player is done from the press box and NOT all journalists are asked..better still its affected by the sponsors. The reality is Match Ratings have always been subjective, FM is a simulation and the match ratings here may be a direct result of the team winning, the player scoring, and players generally fulfilling their roles..and its probably more objective than any match ratings system you can find in the world. I have contended since day one that this is a factor of peoples tactics and how well they have nailed the roles to the proverbial cross. If its perfectly nailed in YOU will produce a DM with high ratings. All it takes is one person to prove that it can be done, and the whole argument that ratings are broken is flawed.

Interesting.

Hadn't really considered that despite the fact that I have achieved pretty decent GK ratings for a GK playing the SK(a) role and scoring goals.

Might have to go back and have a play with defensive midfield roles/instructions. (Think someone was going on about a DLP the other day/week on twitter).

[sirgiorgio]

Stands up and:applause: FINALLY!!!! Match ratings are SUBJECTIVE...yes..there is NO verifiable system ANYWHERE in the WORLD. Squawka ratings are squawka ratings, and when you are watching it on the telly, its the commentators who sit down and hash it out, giving ratings. I know cause thats what I used to do for a living. MOM award for a player is done from the press box and NOT all journalists are asked..better still its affected by the sponsors. The reality is Match Ratings have always been subjective, FM is a simulation and the match ratings here may be a direct result of the team winning, the player scoring, and players generally fulfilling their roles..and its probably more objective than any match ratings system you can find in the world. I have contended since day one that this is a factor of peoples tactics and how well they have nailed the roles to the proverbial cross. If its perfectly nailed in YOU will produce a DM with high ratings. All it takes is one person to prove that it can be done, and the whole argument that ratings are broken is flawed.

I used to do that for a living, too. Well, it was more like a hobby really: I got paid, but unfortunately not to watch Champions League matches...

Anyway, one thing is "match ratings are subjective", one thing is "there's nothing SI can do to improve them".

Goalkeepers. If a 'keeper saves 10 shots and keeps a clean sheet, it's hard not to rate him as the player of the match. Sometimes goalkeepers are praised "just because" they make one key save (not necessarily a penalty, but a very difficult save in a key moment of the match). "That save was as important as a goal", is the inevitable reaction of the press. In FM, it's not rare to see a team having 10 or more shots on target and no goals scored, and the PoM is some random central defender...

Red cards. A straight red, especially very early in the match, is possibly the most stupid thing a player can do. In my book, and not only in mine, it's a 4. In FM, a red card, justified or not, doesn't seem to make any difference.

When a penalty is awarded, it's considered a "mistake leading to goal". But then, in the commentary line, one reads "the penalty did seem harsh", and the referee "made the controversial decision to award a penalty..." So was it a penalty? If it wasn't, a poor match rating is like adding insult to injury.

Also, when a goal is scored, it's not unusual to blame someone in particular, something you seldom see in FM. Some mistakes are clumsier and easier to spot, but as I wrote above you'd expect the game to see a mistake for what it really is. An header from inside the goal area? Blame the goalkeeper. A goal from corner? Blame the defender who was man-marking the scorer (or, in case of zonal marking, the defenders who were closer to the ball). A striker beats the offside trap? 8 times out of 10 it's because a defender didn't keep the line. A "scorcher from distance"? It's either the 'keeper's fault, or a midfielder's, who didn't close down the shooter.

It's football, not rocket science!!!

One thing to consider, though, is that real match ratings are given when the match has ended, while ratings in FM change in real time. This is a very important distinction to make. An old school poacher can play hide-and-seek for the most part of the match just to appear in the right place at the right time to score the easiest (and in general, crucial) goal. If one has to rate his performance in real time, his rating would be 4 / 5 / 4 / 3 / 3 / 3 / 8!!! Defensive/Holding midfielders "suffer" from the same problem, for the opposite reason: their job is often obscure, tactically very important but not "flashy", and more often than not you can "feel" their presence but it's only very late in the game that you realise how important they are for their team. Thus, in general, their ratings are high or low depending on their team's overall performance.

[Kenco]

Agree with the bit about red cards above - I've mentioned this several times over the years but nothing has changed.

[Jimbokav1971]

Red cards. A straight red, especially very early in the match, is possibly the most stupid thing a player can do. In my book, and not only in mine, it's a 4. In FM, a red card, justified or not, doesn't seem to make any difference.

It's football, not rocket science!!!

Right, I've never done this for a living, but let me just give you my tuppence worth to show quite how hard it is to manage something subjective like this.

You have given an example of a player getting a straight red early in the game. You don't say how early, but just for arguments sake, let's say that it was shown on 09:00 exactly into the game.

Now for some reason, (remembering that we are scoring the player out of 10), you have somehow decided to give this player a score of 4. I don't understand that at all. What has the player done to deserve 1, 2 or 3 out of 10 never mind about 4 out of 10? He has played approximately only 1/10 of the game and has managed to get himself sent off. If he hasn't done anything of note, how can he be scored anything more than 1, and more likely actually 0?

Where is this unwritten rule that when rating a player out of a maximum of 10, the lowest rating that can be given is a 4?

If you can only get a score between 4-10, then why not just make it 0-6?

In reality, even poor performances are usually scored at 6+

A really poor performance can often be a 6+ and a really good performance can often be less than 9.

It just seems that lots of ratings are being squashed into 6-9. We actually don't have a 0-10 rating system at all.

Why doesn't a player who is sent off in the opening 10 minutes of a match awarded a 1 or a 0 out of 10?

Why do ratings start at 6 and then move up, (even when a player performs poorly), from there.

4 does indeed seem to be the minimum that you can get.

I have recently had a centre-back who played for 54 minutes, and was apparently responsible for both the goals we conceded.

Despite that, his stats were as follows.

Passes 22 Completed 16 Key 0.

Tackles 3 Won 3 Key 1.

Headers 2 Won 2 Key 0.

Interceptions 7.

Runs 1.

Fouls 1.

Rating 4.9.

That's the lowest rating I have seen in my current save over many many seasons.

[SMV90]

I've just completed a match in which I had three players finish on a rating of 9.0 - I don't think I've seen that before.

Alan Richards (CM/S) got 1 goal and 1 assist, and he was voted MoM.

Sam Wellington (BBM/S) got 2 assists.

Declan Jones (IF/A) got 1 goal.

I know this doesn't contribute much to the topic at hand, but I thought I would share it anyway. Sorry

[sirgiorgio]

Where is this unwritten rule that when rating a player out of a maximum of 10, the lowest rating that can be given is a 4?

In Italy (where I live) the morm is to give ratings from 4 to 8. You can see an odd 3 (a player who's sent off after 9 minutes would be a likely candidate) or a 9 (a player scoring 4 goals, for instance). A poor performance would score 4,5 or 5. An average performance would score 5,5 or 6 (depending on the team's overall performance). A good performance will score 6,5 or 7 or 7,5 (depending on a number of factors). 8 is given to a real standout performance. This is how it works where I live (and also more or less how it worked in the good old days of Championship Manager, IIRC), but it doesn't work like this everywhere. After the infamous 2014 World Cup match between Germany and Brazil (7-1) Marca (Spanish newspaper) chose to assign a score of 1 to every Brazil player, and 10 to every Germany player.

Now, these are norms, or conventions, or unwritten rules. Subjectivity is another matter, which has very little to do with Football Manager, and with the examples I gave above. A great save is a great save, period. A straight red is a straight red. A defensive mistake is a defensive mistake. A goal is a goal. These are facts, that everyone can see. When someone scores, his match rating increases significantly. How much significantly it should increase is subjective, and I’m ok with that. When a player is sent off, his match rating should decrease significantly, and this just doesn’t happen. I had a guy sent off after literally 2 minutes after a stupid, useless, gratuitous foul, and his rating was -.- , as in “did not play enough to be rated”. He did play, the idiot! You want to rate him 1, 2 or 5? That’s not important, as along as he gets the lowest possible rating according to your own (more or less written) rules. If the rule in FM is 5.5, that’s ok, if the rule is “let’s pretend it didn’t happen”, that’s not ok.