Friday, January 27, 2012

Gameweek 23 Captain Rankings

P90 Points per 90 minutes at home/away based on the player's fixture QG Number of quality games for the season (6 or more points) BG Number of bad games for the season (3 points or less) % Percentage of team goals scored or assisted SD Standard deviation of player scores (risk) Rating Percentage of team goals accounted for multiplied by expected team goals for the week.



A couple of readers last week said it might be helpful to get a bit of context for the above stats, so below are a few of highlights for the week:

Everything points to Van Persie for the week, so while I generally stress that the 'Rating' column is not a pure ranking, he would indeed be my number one pick this week. No one can match his combination of upside and consistency. Given the home fixtures enjoyed by Spurs and Man Utd it's rare I will categorically take a road team but Bolton aren't much better at home than they are away, making Van Persie a very good play this week.

Rooney will gain a lot of attention this week, and rightly so, though I'm slightly troubled by his recent performances. His home P90 is very impressive (7.3) but if you look at just the last six games at Old Trafford that number drops to just 5.0, largely due to the removal of his massive GW3 haul. Now I don't want to play around with numbers too much just to fit the narrative but he has clearly struggled for consistency (with his production if not his underlying performance stats which remain strong). Given my suspicion regarding 'form', I'd suggest Ronney is a regression candidate and should start taking his chances more frequently going forward, but I'm not sure I'd be willing to bet my captain's armband on it. For the record, I love Valencia as an asset to own but I'm not sure I'm ready to throw him too much captain consideration just yet.

It's rare that the City boys aren't considered great plays in a given week but their trip to Everton is tricky compared to some of the other fixtures around this week and they go into the game in dangerous, but inconsistent, form. Aguero is the obvious pick but with Balotelli out, Dzeko is looking like a very interesting option, especially if you put any stock in his home performances to date (small sample size alert!). Silva's main appeal is his consistency, as evidenced by his SD of 3.4. None of the other elite options have a floor as high as Silva, who seems to notch at least some points every time he takes the field (his 7 BG in 22 appearances is also the best mark among elite options). There's an argument to be made for always captaining a player of Aguero's quality but if you're following the stats, it's likely you'll have a better option available this week.

Adebayor, Van der Vaart and particularly Bale will get a lot of attention this week too, and with Wigan coming to town the optimism is well placed. Spurs' fixture is comparable to Arsenal's, arguably a touch better given that teams tend to hit their potential more at home than on the road. The problem is that the Spurs trio are all aiming for a piece of the same pie. All three have managed their fair share of quality games this year, and one of them has managed a QG in every game for Spurs bar three. The problem is that only six games have seen two of them notch 6 or more points and this hasn't happened since GW12. By captaining one from this group then, you are really taking an educated guess as to which of the trio will enjoy the success this week. It may seem obvious that it would be Bale but don't forget that both Adebayor (GW11-18) and Van der Vaart (GW6-10) have both enjoyed their own spell of fantasy dominance this year. Of the three I would take the safe play with Bale, but I am actually considering an official 'recommend' stamp for Van der Vaart for the next few weeks, who essentially brings everything Bale does, along with an increased threat from set pieces.

Dempsey is sure to get a look from many this week after his hattrick last week, and while I don't put much stock in prior results, his performances of late have screamed of a player ready to start scoring plenty of goals. Fulham's ceiling as a team will never be able to match the big boys so you're hoping Dempsey's increased importance in that team equalises that fact. Those 12 BGs terrify me and were the main reason I passed on him as captain last week. That said, if you look at how Dempsey has done against weaker opponents at home, the results have generally been very good so there is at least some basis for believing he will notch one of those 8-10 point games. Risky, but I see the logic.

The pleasing thing about the 'rating' category this week is that it is throwing up a good mix of familiar and obvious names, along with a couple of wildcards who people may not be thinking too much about. Bellamy and Zamora have been extremely successful when they've played this year and both should get another chance for success this week. You have to caution against the very small sample size these numbers are based on, but they've taken their chances so far so you wouldn't be amazed to see at least a large portion of this success continue. I'd still probably rate both as third forwards though, unless you decide to go huge in midfield. They are only captain options if you're starting to get desperate to start making ground with opponents whose team is largely similar to yours.

19 comments:

David said...

Hi Chris! Thanks for the excellent info you put out every week! I am loving the tables by the way. What are your views on my team for this week:

Vorm
Kompany, Riise, BAE
Silva, Dempsey, Bale(VC), Sess, McClean
Aguero, RVP(C)

Mignolet / Fletcher, Williamson, Huth

Going with stats from your tables I have chosen Vorm over Mignolet. You have the same keeper combo, which one will you be playing?

Stats say play Fletcher but stats also say Liverpool are one of the best away defenses regardless of what happened last week. My gut says play McClean.

What do you think of my team overall?

Thanks again!

JS_shadow said...

Chris! Thank you so much for your efforts!

The table and your "GREAT" comments combined makes the your captain rankings even better!!

Can't wait to see your next week's posting!!!!

Rutwik said...

Hey Chris great work.. can you write one on how the price of a player increases or how the FPL people work it out. Coz i think the extra cash helps a lot. I have seen teams with 110 mil value. it helps a lot.
Thanks :)

DLL said...

@ Rutwik

You can use this site to tell you when players are due to rise http://www.fiso.co.uk/crackthecode.php
Price increases are just based on NTI (Net Transfers In). Drops are a bit more complicated I think, the number of NTO needed to result in a drop depends on how many people own the player.

vanilla said...

Future column idea (for slow days during international/UCL fixtures): variables that effect how useful an astute 4pt extra transfer can be. Yes, I've had some personal success in the transfer market this season.
Since I have been a steady 60-70 points back right from the go (I did not own Rooney or Ashley Young for the first month, while my competitors had them captained), I have found it easier to legitimize several 2-transfer weeks.
I took a 4pt hit to both diversify my team (remember that GW1 freebie wildcard bungle? it was instant death to shrewd teams) and to steal a march on rising stars.
As a result, I am ~4.3M higher than my league-leader in team budget. This means that my "subs" are not Gower/Moses/Kelly/Morison/Graham, but more like Fletcher/Enrique/Senderos/N.Taylor.
This is a boon for my team, because as someone mentioned in the comments on your previous post, this has been the Year Of The Rotation Dragon.
As a result, my lineup decision each week is not "which matchups do I avoid" (needing to take a punt on someone like Pilkington) but rather "which matchups are strongest" and "Do I really want to sit Fletcher at home against Liverpool so I can play all 5 of my midfielders"
The end result remains to be seen, but I've been able to shave the 80 point gap down to 30 over the course of the last 4 gameweeks.
Sorry about the long winded post.

stooshermadness said...

Chris - great post, these new Captain's Ranking posts are fantastic. I wonder if anyone else has concerns about RvP; the Guardian Squad Sheets (I know, not all that accurate) has RvP starting on Sunday in the FA Cup. Then Arsenal have Bolton next Wed and Blackburn next Sat. 3 games in 6 days. If, indeed, he starts Sunday I'm a little less confident about the armband for Bolton. Anyone seen anything on Wenger's thinking ?

chemikills said...

New captain rankings are really good, love it. I'll certainly be checking them weekly rather than checking your site here and there ;)

I was attempting to be retrospective and point out how well your rankings did. Unfortunately your previous post has updated to the new rankings so I can't ^^
But from memory it was very good :)

chemikills said...

I was looking for a source for "points per game" stats to help compare some of those players that just don't get to play the full 90.

I found this:
http://userscripts.org/scripts/show/34047

It's super useful, figured some people on here might like to check it out.
It's a script that adds a few things to your FPL page :)
I'm still playing with it so can't detail everything but it's very nice

Mr Greedy said...

Loving the captain rankings, this is a great addition to the blog. I'm just a bit worried about how useful the standard deviation statistic is. Do we necessarily want a low standard deviation from our captain? A player returning scores of 16, 6, 2, 13, 18, 2, 2 ... will have a very high S.D but be good captain material, especially if those high scores were from predictable easy home games. A player with 2, 2, 1, 2, 2, 2 ... is obv going to have a very low sd but would not be captain material. Even someone like Silva I would class his consistent 5, 6, 5, 6 style as a reason to not captain him, although he is a good VC.

Mitchell Stirling said...

Hi Chris,

I've used my wildcard and took on your advice on squandering a 6m midfielder each week on the bench. My current team for week 23 is

Vorm / Skrtel, Evans J, Dawson / Dempsey, Bale (VC), McClean, Gerrard / Fletcher, Van Persie (C), Rooney

Bench: Mingolet, Richards, Moses, Caulker.

My reasoning for benching Richards is his poor record away from home and the logic that King is more likely to be rested for Wigan at home allowing Dawson to play. If Dawson doesn't play, being a defender he is less likely to get junk time in last 20 mins or so.

For the rest of the season I have 0.6m in the bank, no wildcard and not planning a transfer this week in case there is some activity before midnight Wednesday (not expecting much) I'm c65K overall, top or there abouts in my smaller mini leagues but my main goals are to top 2k points and catch someone who is 79 points ahead of me (to win money)

I have a habit of using 4 points most weeks and have a 50% strike rate of it being worth using.

I'm now convinced that 3-4-3 is the best formation to use unless you have strong feeling that four clean sheets are very likely at the back.

Players on my radar are Lindegaard in goal, Clark, Shawcross, Simpson, O'Shea and Rangel in def, Valencia, Mata and Petrov in midf and Zamora up front. While I don't want to ditch too many high scoring players, I do need a couple of differentials in the team to try and close some gaps, without getting caught myself.

Anyone have any thoughts.

Chris Glover said...

David - The team looks great overall. There isn't a single player I would be targetting to sell, though my general concerns about a lack of differentiation may apply depending on your league situation.

As for this week, the stats say play Vorm but most people's gut would lead them to take Sunderland given their opponent. If you believe the team is more important, take Vorm, if you believe the opponent is the key to clean sheets, take Mig.

I'd play Fletcher on the basis that I believe he will contribute more consistently than McClean over a large sample size, but McClean is a useful play this week too. Liverpool can be good at the back and Wolves aren;t elite, though the generally get a goal or two at home and you like Fletch to be involved in most things they do.

Chris Glover said...

Thanks JS-Shadow - the tables are still a work in progress and as Chemikilis points out below, it's not ideal that the old posts now update, though it makes my life ALOT easier by having them auto update.

Chris Glover said...

Rutwik - a few people have asked about team value before and honestly it's something I haven't had much success with in the past. I'm going to add a quick post on this after I answer these questions, as I have a few things to say on this issue.

Chris Glover said...

Vanilla - it's a great question and one I've thought about trying to answer before. I've never been able to really come up with anything better than 'you need the new player to outscore his replacement by 4 points' to break even. Given that few players average 6 points a game, you;re left with a situation where the only obvious 4 point situations are where:
(a) you only have 10 players and thus your new arrival would only need to score 4 points to break even (with everything else being upside), or
(b) double gameweeks, where the new transfer is all but guaranteed 4 points and thus only needs to earn 2 more to help you break even.

Now I have the new '% created' stat, I might try and make this analysis more scientific. If a player gets an assist or scores a goal, he is very likely to hit the magic 6 point line, so we should be able to work out a point at which the risk might be worth it.

Chris Glover said...

Stoosher - it looks like Van Persie will play tomorrow, and hence, you're right, at some point he's going to get a game off in the league, given the Milan game is now just over two weeks away. You'd point to Blackburn at home as a more likely fixture for him to take a break but he's going to be a risk for the next few weeks until he finally sits. It's a consideration for sure, but I've kind of given up trying to predict manager decisions, unless there's strong evidence.

Chris Glover said...

Chemikills - it's a downside that the rankings now update every week, and thinking about it, if I were to save the tables as a new filename each week, I might be able to avoid this. If anyone uses Tableau let me know, or else I will keep experimenting to get this right.

The script looks interesting. Are there any stats you or anyone else would like added to the player data tables?

Chris Glover said...

Mr Greedy - it's a good point and one I haven't elaborated on yet as I was planning a piece for FFS on the very issue. I will touch on it here though. Standard deviation in itself isn't that useful, but as supplement to the other data I think it has some value. You're right that a player who scored 12-2-6-4 would have more than value a 2-3-2-2 player, but how about a player who scored 6-5-6-7 ? I would definitely want to own the more consistent player, and if playing from a strong position would want to captain him too (if playing catch up you may opt for upside).

You're right that as stats go it's not the best, but I thought it had just enough value to warrant inclusion.

Chris Glover said...

Mitchell - I think you're on a good course. The team looks very good and you've identified a good list of differential ideas to change things up going forward. I really like Zamora and Mata as high upside differentials.

As for the best formation, the success of the big name forwards is making three forwards look very attractive while the struggles of the 'big' defenses makes 3 at the back look like a good way to go. I'd therefore agree that 3-4-3 is often going to be your best bet. I don't generally favour one and take each week as it comes but, without looking, I'd guess I've gone 3-4-3 or 4-4-2 most often (I had a budget forward in Graham until my wildcard).

Sqiar BI said...

Thanks a lot for sharing this with all folks you really recognise what you are talking about! In this complex environment business need to present there company data in meaningful way.Sqiar (http://www.sqiar.com/solutions/technology/tableau/) which is in UK,provide services like Tableau and Data Warehousing etc .In these services sqiar experts convert company data into meaningful way.