Stats on shots, chances, on target shots

[ggwood]

I recorded some information on shots in FM17. I've done this before, I think it was with FM'15. Things have changed a bit. It is more likely to get a chance (clear cut or half chance) from a header. Further, if your player is one-on-one with the keeper (which usually would be a clear cut chance), but the keeper is blocking most all of the frame of goal, it will not result in a chance as often. Also, you can get a half-chance from a free kick that you miss, if the free kick is very close to the edge of the box. It used to be you would only get a half chance if you scored from that free kick, so it skewed the half chance stats.

Here is what I have added up thus far:

	SHOTS	ONt	CCC	HC	GCC	GHC	GOT
	820	367	92	121	35	22	47

From about 38 games from one season in the English Championship with my first season after taking over Blackburn Rovers (I finally got here in the year 2023 or something). So: 820 shots, 367 of which are on-target. 92 clear cut chances, from which 35 were scored. 121 half-chances, from which 22 were scored. 47 goals from on-target shots, which were not recorded as chances.

About 38% of clear-cut chances result in a goal.

About 18% of half-chances result in a goal.

About 13% of shots, which are not chances, result in goals.

About 28% of on-target shots, which are not chances, result in goals.

Key takeaway: "clear cut" chances are not a sure thing. On-target shots are a lot better then I thought they would be. Half-chances are not as great as I thought they would be.

I fully expect the percentages to change when I get more data. 22 goals from half-chances: treat that as plus or minus at least four, which is a huge uncertainty, so the goal chance from half-chance could easily be 15% to 22%, and that is just one example. (I can explain, this is from Poisson uncertainty, which is the best we can possibly hope for).

I'm now recording for each chance what happens: is it on or off target, and is it a goal (or does it result in an own goal, so off target, but a goal results). Further, I'm breaking it down by type of shot into: kick, header, free kick and penalty. Kicks and headers can be half or clear cut chances, while free kicks can be half chances, and penalties are clear cut chances. Although the data above I do keep broken down between me and each, individual opponent I have faced, honestly, the percentages are about the same, so in this new, more extensive data set, I'm lumping me and my opponents together.

 

[SmoothJitsu]

The way CCC's and HC's are awarded in this game is completely wack. Like most of the game. haven't enjoyed FM at all this year.

 

[vanTiggelen]

Yep, clear-cuts have all been all over the shop. That is, insofar you could say it. Given it's a wholly subjective stats, it's not defined anywhere in the game. The madness of which is right in the first post actually, as in the admirable attempt to make sense of this, ggwood works backwards from looking at the play and tries to decipher what may factor in. And as he has found, on some iteration header finishes may count, on others less (from my experience they are out of the contest still, even if the goal is empty and the header made under no pressure from point blank rage --- I remember there was debate about this seasons ago).

Here is what I have added up thus far:

	SHOTS	ONt	CCC	HC	GCC	GHC	GOT
	820	367	92	121	35	22	47

From about 38 games from one season in the English Championship with my first season after taking over Blackburn Rovers (I finally got here in the year 2023 or something). So: 820 shots, 367 of which are on-target. 92 clear cut chances, from which 35 were scored. 121 half-chances, from which 22 were scored. 47 goals from on-target shots, which were not recorded as chances.

About 38% of clear-cut chances result in a goal.

About 18% of half-chances result in a goal.

About 13% of shots, which are not chances, result in goals.

About 28% of on-target shots, which are not chances, result in goals.

Key takeaway: "clear cut" chances are not a sure thing. On-target shots are a lot better then I thought they would be. Half-chances are not as great as I thought they would be.

Which brings us to this: Opta clear-cuts are converted at about 33% rates. Given that none if this is stated anywhere in the game, and that the target ain't Fifa, these are some real-life "big chance" conversion stats as per Opta, typically genuinelly one on ones. Arsenal's Premier League title hopes hit by poor finishing | Football News | Sky Sports Given that from open play there is barely anything even touching a penalty outside of tap-ins into empty nets perhaps, you'd expect that to be in there in the game somewhat roughly. And it is, check every frustration post there has ever been. If you're of the opinion the keeper isn't the advantage from open play in one on ones most of the time, the game will never please you, as that is SI's stance. It is backed up by research.

SOT percentages are also pretty close to real football, about 1 in 3 to 1 in 4. Careful as the game has never made a distinction between shots off set pieces and open play though. The set piece ones, unless you have found an exploit, are always harder to convert, and can pile up a plenty. All for tactically reasons, typically. There is very little coincidence that if you look into final match stats screenshots posted in those frustration threads, you often see a very high amount of corners for the "dominating" team. Unfortunately the post-match feedback has never picked up on this, and rather makes out a side just being "wasteful" rather than what is really happening: It struggling to break down a defense much at all.

Great and interesting efford this! 8-|

 

[ggwood]

I think you can control your shots on target percentage by "working the ball into the box" or directing players to shoot less. Basically, you can have more, low quality, shots, or fewer, higher quality ones.

The thing you need to analyze this is the "long shots", which I just cannot do. It is just too much to hold in my head all at once. When I did try to do this, (in FM'15) I concluded something like 2% of long shots resulted in goals, but that number was so small, it could have been double.

I cannot find my old data from FM'15, but my recollection is that clear-cut chances had a higher rate of success, maybe 45%?

There was a goal that went in like this: a shot comes in hard, a defender tries to clear it, but a second attacker is right in front of the defender, both of whom are very close to the goal. So the defender kicks it straight into the shin/knee of the attacker, where it deflects straight into the goal. I was expecting this to count as a chance of some kind, but it is simply an on-target shot, goal.

Chances from headers are basically a complete mystery to me. Generally, you have to be very close to goal for your header to be recorded as a chance of any kind, and generally for it to be clear cut, you have to a pretty good "view" of the goal, meaning it cannot be totally blocked by defenders or the keeper. Thus far, I've only got four clear-cut chances from headers (one goal, and two on-target non-goals) compared to ten half-chances from headers (two goals, five on-target non-goals). These numbers are so small, I cannot draw any conclusions from them at all, but I started really late doing this. Of the 38 games where I kept track, I think I only did like ten games where I kept detailed track of each chance.

I got promoted to the premier league now, so I'm probably going to create a whole lot fewer chances. Having done this before, I can pretty much assure you that the percentages won't change.

 

[Lojo87]

I find that often half chances are even more dangerous than some clear cut chances

 

[ggwood]

Do not only focus on clear-cut chances. Most goals are not scored from clear-cut chances. In my large data set, only 35.6% of all goals come from CCC. 22.7% from half chances. 39.1% are from on target shots. About 2.7% of all goals are own-goals. Due to rounding, they add to 100.1%.

Next, your chance of scoring from a clear-cut chance depends on what kind of CCC it is. If it is a penalty, about 8 in 10 are scored, BUT I find this is lower in top leagues.

Only about 34% of clear-cut chances kicked from open play result in goals. I have evidence headers which show up as CCC are somewhat less successful. I don't have enough events to report yet (only 18 events, only 5 goals from this, but I cannot say anything from that).

About 25% of half-chances kicked from open play result in goals. Half chance headers are worse. I found only 4 headed half chance goals from 39 half-chance headers, which is 10%! Now, you can also get a half chance from a free kick. I only have one goal from eight - I cannot say much from that. Note: in this time, I'm sure I've seen a dozen free kick goals - they just were not recorded as half-chances.

I think from this you'll see if you kick, from open play, half-chances are more similar to clear-cut chances then I reported in the first post above (originally, 38% to 18%, but when we only look at kicks from open play, 34% vs. 25%).

Note on headers: I cannot look at a header and tell what it will be recorded as. With kicks, I generally have some idea. This is still better then FM'15 where virtually no headers were recorded as chances.

 

[Death Ball]

I've just been directed to this thread, that I somehow missed. Great thread and contributions, thank you all. I'm so happy somebody got around to do this and contribute these awesome posts. :wub:

 

[ggwood]

A few notes on chances:

1. You cannot get a "chance" unless your player shoots the ball.
2. The factors that seem to determine if something is a "chance" are: (1) proximity to goal, (2) view of the goal (is anyone in the way?), and (3) time on the ball (are defenders close?).

Stats (shots, chances, etc) are buffered for a short time, and all events which occur during that brief time are reported together. Watch this video, just before the six minute mark: https://www.youtube.com/watch?v=sBUWrWDaWUs where a long cross comes down, the striker heads the ball, goalkeeper blocks, the striker taps the ball into the net. Together, the stats update as: +1 clear-cut chance (CCC), +1 half-chance, +2 on-target shots, +2 shots. Now: we have to guess which shot was considered a CCC and which was a half chance. I would score this as the header was the half-chance, and the kick was clear-cut. I think you will agree with this. However, I run into times when the two shots could go either way - this blurs the stats you see above.

My team got promoted, and I have to play defensively to get any results and stay up, so now I'm seeing very, very few chances per game. Which means it is taking a long time to accumulate new stats. I'm getting ready to update soon.

 

[ggwood]

Stats on Shots, CCC, On-Target, Off-Target, etc.

Not all clear cut chances are created equal: penalties are better. Headers are worse.
For half-chances, headers are somewhat worse than kicked half-chances. Free kicks are sometimes recorded as half chances. I have so few, it is hard to say. I guess they are about the same as normal half chances, but they are just very rare.

I don't think the numbers have moved much since the last time I posted them. I'm working on a bigger project which may be harder to understand how it works, but the results may be of interest.

Chances and their Results:
Kicks:
Clear Cut Chances Kicked:
Goals	On Target (no)	Off Target	N Events	% goal	%on Target
92	76	59	227	0.4052863436	0.7400881057
Half-Chances Kicked:
Goals
54	111	107	272	0.1985294118	0.6066176471
Free Kicks: Half Chances
Goals	On Target	Off Target		% are Goals:	%on-Target:
3	0	11	14	0.2142857143	0.2142857143
Penalties: CCC
Goals	On Target	Off Target
14	2	2	18	0.7777777778	0.8888888889
Headers
Clear Cut Chances: Headers:
Goals	On Taret	Off Target
8	11	15	34	0.2352941176	0.5588235294
Half-Chances: Headers
Goals	On Target	Off Target
10	26	33	69	0.1449275362	0.5217391304
CCC TOTALS:
Goals	on T	Off T
114	89	76	279	0.4086021505	0.7275985663
Half-Ch TOTALS
Goals	on T	Off T
67	137	151	355	0.1887323944	0.5746478873