Tactical Ramble related to Wood's departure and replacement.

[SuperJimmy]

Another factor to consider is penalties.

Last season, Wood scored 5 in the league, 1 in the EFL cup. If another player had taken and scored those, Wood would've had 22 league goals. That's not irreplaceable by a half-decent striker and more attacking contribution from midfield.

Good point as well, makes a huge difference that, 18.5% of Wood's goals. Moves the goalposts on what you are really trying to replace.

[SuperJimmy]

Good work, well presented, an interesting read. Statistical analysis of football is just so difficult - there are so many variables!

My gut feelings for the season to come are:

1) Regardless of who we have to replace Wood, he won't scored 30 goals.
2) Our midfield will contribute significantly more goals this season.

Whether that means we will be more successful in terms of league position at the end of the season, who knows? But I'm certainly more optimistic about our chances now, than I was this time last year.

This is pretty much the crux of the whole thing, I agree 100% with you on the first point and that he wouldn't have scored 30.

I think that the midfield are capable of scoring more but I don't KNOW if they will. We need a big jump.

Time will tell

[SuperJimmy]

Great to see some interest for this kind of thing. Won't be a regular feature in such a long format but as I said I was having a look anyway so thought I'd share.

General consensus is we need more goals, from more people!

[SiMamu]

The other side of this is if we sign a striker that isn't a Target Man, for comparison I'm going to look at when Ross McCormack was at Leeds. In the 2011/12 season where we achieved our highest goal tally of 65 since 2010, McCormack only scored 18 goals, unfortunately I can't look back as far to see what proportion of Leeds' chances fell to him. The other notable goal scorers were Snodgrass at 12, Becchio at 10.

Obviously this system was 4-4-2, so two strikers sharing the load of goals, but that's exactly the question I'm after, if we think we need more goals, is it as simple as personnel or does the system need changing.

Worth noting that in the 2013/14 season with the same system I believe, McCormack scored 28 goals, Matt Smith got 12, so the striker pairing scored 12 more goals than the 2011/12 pairing of McCormack and Becchio, however the team finished 7 goals less than the 2011/12 season.

Did losing Snodgrass make all the difference?

Just a note to factor in, Becchio left us in January of that season I'm sure so it could be argued that 2011/12's pairing was just as prolific as McCormack/Smith.

Excellent job with the stats though, it's an interesting read and i feel it goes to show just how good a season Wood had.

Good point and one I missed! Definitely seems that having two strikers share's the load better. Wonder why it's now so out of fashion.

Teams want an additional midfielder to try and take control of things more and to help with building attacks at a slower pace, playing it along the ground - with more midfielders helping to make this a smoother transition.

[kk white]

Lets be honest folks, while we love to surmise on recent seasons records, deep down the truth is that there are so many factors at play that we just can't predict/evaluate what is required to score/replace goals. Statistics in their true sense show that spikes and dips regularly appear in data without obvious justification. Fact, whether we choose to accept it or not. The more data we choose to evaluate, the more obvious pattern we tend to see. Statistics in their right context, with a long enough history of data, are very powerful. Otherwise there would be millions of people out of work and hundreds of major companies out of business. Last season could be a spike in the 15 year career of Wood (VERY possible in my opinion) which coincided with a dip in midfielders scoring. Most managers will eventually admit that luck plays a vital part in success. Bad luck has seen many a good manager's demise. Luck = statistics = all things eventually being equal.

I genuinely don't mean this reply as an argument to SuperJimmy's post. Loved it and it had me thinking which is what I love about this forum. I agree with everything Jimmy said because the statistics are there to back it up, but we need a little luck to combat the norm every now and again. Please let it be Orta, Radz, TC and their wonderful signings this year. MOT.

[Deebo]

Many thanks for that considered and thought provoking post SuperJimmy. Enjoyed reading that....

[johnh]

The other side of this is if we sign a striker that isn't a Target Man, for comparison I'm going to look at when Ross McCormack was at Leeds. In the 2011/12 season where we achieved our highest goal tally of 65 since 2010, McCormack only scored 18 goals, unfortunately I can't look back as far to see what proportion of Leeds' chances fell to him. The other notable goal scorers were Snodgrass at 12, Becchio at 10.

Obviously this system was 4-4-2, so two strikers sharing the load of goals, but that's exactly the question I'm after, if we think we need more goals, is it as simple as personnel or does the system need changing.

Worth noting that in the 2013/14 season with the same system I believe, McCormack scored 28 goals, Matt Smith got 12, so the striker pairing scored 12 more goals than the 2011/12 pairing of McCormack and Becchio, however the team finished 7 goals less than the 2011/12 season.

Did losing Snodgrass make all the difference?

Just a note to factor in, Becchio left us in January of that season I'm sure so it could be argued that 2011/12's pairing was just as prolific as McCormack/Smith.

Excellent job with the stats though, it's an interesting read and i feel it goes to show just how good a season Wood had.

Good point and one I missed! Definitely seems that having two strikers share's the load better. Wonder why it's now so out of fashion.

I think there are two reasons for twin strikers going out of fashion. The obvious one is that it provides the opportunity for an extra mid-fielder or a back five. The other reason is the improved fitness of the lone striker. Look at how Wood improved in this respect from the season before last to last season. Everton did it with Lukaku. The reason the lone striker gets a high proportion of the goals tally is that he is usually the only player in the box when the ball's there. However, I'm not a fan of the lone striker. I think we have to change tactics to have more people getting on the score sheet.

[SuperJimmy]

Lets be honest folks, while we love to surmise on recent seasons records, deep down the truth is that there are so many factors at play that we just can't predict/evaluate what is required to score/replace goals. Statistics in their true sense show that spikes and dips regularly appear in data without obvious justification. Fact, whether we choose to accept it or not. The more data we choose to evaluate, the more obvious pattern we tend to see. Statistics in their right context, with a long enough history of data, are very powerful. Otherwise there would be millions of people out of work and hundreds of major companies out of business. Last season could be a spike in the 15 year career of Wood (VERY possible in my opinion) which coincided with a dip in midfielders scoring. Most managers will eventually admit that luck plays a vital part in success. Bad luck has seen many a good manager's demise. Luck = statistics = all things eventually being equal.

I genuinely don't mean this reply as an argument to SuperJimmy's post. Loved it and it had me thinking which is what I love about this forum. I agree with everything Jimmy said because the statistics are there to back it up, but we need a little luck to combat the norm every now and again. Please let it be Orta, Radz, TC and their wonderful signings this year. MOT.

I agree entirely with the majority of this post, bar a couple of things.

I wasn't specifically talking about midfielders scoring, more just the distribution of goals between players regardless of where they play. We just need more players getting 10+ rather than a striker getting 25.

This may have been what you meant, however if it was then I think looking back to 2010. Between the 2012/13 season and last season we consistently have 1-2 higher scorers, and then a severe drop to the next group. This excludes 14/15 where the distribution was better but the team only scored about 50 goals because the top scorers were just so low, Antennuci was top with 10 I think, or around that

4 seasons doesn't sound like a big dataset, but looking at goals per season is an identical analysis as looking at goals per game, you're just aggregating the data points,for example if a player scores 23 goals a season then by analysing every game you would have 0.5 goals a game, and a dataset of 46 games. So over 4 seasons we are considering 184 games, which is a big enough data set to have explanatory power.

As was pointed out earlier 2010/11 and 2011/12 seasons were much different, we had much smoother goal distribution, and we haven't beaten the goal totals of those seasons since then. Two seasons is 92 games, which is still large enough to be of statistical power.

These aren't revolutionary findings, anyone who watches football, or even people that don't know football can tell that more people scoring more goals = more team goals. BUT it does seem to show that having one man getting 27 or 28 goals a season as Wood did in 2016 and McCormack did in one of the last 4 seasons, doesn't translate across to overall team goals. Take Fulham and Leeds last year, our top scorer had 28, there's had 12. They scored 85 team goals, we scored 61 (I'm going from memory from my post so if the numbers are slightly different then forgive me). So despite a 14 goal difference from the top scorers they scored 24 more goals than us.

And this is where I most disagree with you is where you say luck = statistics. Statistics can tell you when luck played a part, such as Wood scoring 27 last season being an outlier, and then you can use statistics to make sure you aren't relying on luck. My analysis on the goals, although very basic as I don't have the time to model it with tons of variables on statistical software, nor do I actually believe would have that much more explanatory power because there is too much variance associated with each variable, I think showed that, without anyone having a break out season that goes against historical trends I.e what Wood did last season, we need more players scoring.

That was designed to start debate as to how we go about replacing them, in my opinion something would have to be done tactically, rather than throwing the same players in the same tactic and hoping they suddenly score more.

I wasn't trying to predict the exact amount very much a start point for debate.

[mtsi5630]

Great post Jimmy, it got me thinking back to the movie Moneyball. Even though it was a baseball movie, the statistics were boiled down to one crucial factor: the players that make it to first base most consistently should be selected.

The factor we should be using as a similar heuristic is chances created. I firmly believe our lack of goals stemmed from a lack of chance creation. Create more chances and more goals will eventuate. We are definitely better placed in this department now compared to last season.

[kk white]

And this is where I most disagree with you is where you say luck = statistics.

Hi Jimmy, I probably typed that incorrectly. My post was basically saying that we (and any team that wants success) eventually need a little luck on our side this year to go with all the good work being carried out and I hope that TC and Radz's team happen to bring some with them. I used the equals signs to say that a bit of everything is required. Probably should have used a + sign. My bad, that's what happens when I attempt late night replies after a 15 hour shift..... And a bevvie or two
Keep up the good posts, really enjoying everyone's take on it. MOT.