Opinion Questions For Rowsus

Rowsus · 2 Mar 2015

Hey Bobbie,
Priddus turns 30 just before the season starts, and his history reads:
20/104, 19/109, 22/114, 20/110 (19/115), 21/101 (20/106), 22/113
The blue scores are what we get, when we remove his 2 concussion affected scores in 2012 and 2013.
Sloane turns 25 just before the season starts, and has returned:
18/88.3, 21/106, 21/107, 22/115

To me it looks like that Priddis is a bankable commodity, that will in all likelihood return something like 20-21 games somewhere in the 106-114, but most likely in the 110-112 range. Sloane is a more "dynamic" commodity, that could return anything in the 105-120 from around 21 games. It is one of those strange situations. I believe Sloane is more likely to be a 115+ player than Priddis, but Priddis is more likely to be a 110+ player than Sloane. So the question that needs to be answered is, in round terms, would you rather bank 110, or roll the dice on a potential 115, that is just as likely to crapout on 106? There in lies your answer.

It does interestingly raise a side question, how do players go in SC the year after their Brownlow win?
[table="width: 580, class: grid, align: center"]
[tr]
[td]Year[/td]
[td]Winner[/td]
[td]SC in win[/td]
[td]Sc year later[/td]
[/tr]
[tr]
[td]2005[/td]
[td]Cousins[/td]
[td]21/124.3[/td]
[td]18/106.6[/td]
[/tr]
[tr]
[td]2006[/td]
[td]Goodes[/td]
[td]22/120.1[/td]
[td]22/104.9[/td]
[/tr]
[tr]
[td]2007[/td]
[td]Bartel[/td]
[td]20/125.8[/td]
[td]22/116.0[/td]
[/tr]
[tr]
[td]2008[/td]
[td]Cooney[/td]
[td]22/106.9[/td]
[td]20/91.2[/td]
[/tr]
[tr]
[td]2009[/td]
[td]Ablett[/td]
[td]19/140.1[/td]
[td]21/129.9[/td]
[/tr]
[tr]
[td]2010[/td]
[td]Judd[/td]
[td]19/118.5[/td]
[td]22/115.5[/td]
[/tr]
[tr]
[td]2011[/td]
[td]Swan[/td]
[td]21/122.0[/td]
[td]18/126.2[/td]
[/tr]
[tr]
[td]2012[/td]
[td]Watson[/td]
[td]22/121.0[/td]
[td]19/111.5[/td]
[/tr]
[tr]
[td]2013[/td]
[td]Ablett[/td]
[td]21/128.9[/td]
[td]15/136.7[/td]
[/tr]
[tr]
[td]2014[/td]
[td]Priddis[/td]
[td]22/112.7[/td]
[td]??/???.?[/td]
[/tr]
[/table]

Only the best 2 SC/Brownlow players in Ablett and Swan have avoided a significant drop the year after their Brownlow win. The average drop is around 8/game. This could be from a number of things, but extra attention seems the most obvious one.
Taking that into account, and the fact that Priddis is getting older, if I was forced to choose between the two, i'd probably go Sloane.
Good luck

Bobbie · 3 Mar 2015

Wow thanks for the comprehensive reply! I still find it hard to choose especially with Priddis' consistency. But Sloane's bye works out better for my structure so perhaps that will be the deciding factor. Thanks again - you have certainly given me something to think about.

Zim^zuM · 3 Mar 2015

Rowsus said:

Not to mention the intensity would be different. A gun will naturally stay in first gear to avoid what happened to libba.

WandP · 3 Mar 2015

Hey Rowsus! I've been tinkering with my team and have about $100k left to spend and I was considering upgrading one of my fwd premos to Robbie Gray. Then I got to thinking about a scoring change for this year that I vaguely remember whereby scaled bonus points for matchwinning goals/contested marks etc were going to be reduced. How much impact do you think this change will have specifically on Robbie Gray and in more broad terms? Is it something that you reckon could be quantified? Just thinking aloud here, but how valid would it be to use averages up to 3 qtr time to effectively disregard the sometimes ridiculous scaling in the last qtr?

Please note, I am in NO way asking you to pull those stats because that would be a ridiculous amount of work (if the qtr by qtr stats are even captured anywhere?) but more asking what you think of the logic?

fodzilla · 3 Mar 2015

Scaling isn't necessarily greater in Q4 of a tight game than, say, Q1 of a blowout if that's where the damage was done. It's just that you notice it when it all comes at the end, rather than being slowly adjusted up through the game as a lead fails to be reeled in.

Rowsus · 3 Mar 2015

WandP said:

Hey WandP,
I think we can over analyse sometimes. From what I recall, the loading factor for this season is half what it was last season. Port Adelaide were behind, or less than 2 goals in front at 3/4 time in 8 of their wins. Let's imagine that Gray benifitted by say 160 points in the "loading zone" across those 8 games. Now the factor has been halved, he would have only benefitted by 80, so it would drop his average by around 4/game, if those figures were accurate. My gut feel tells me the real adjustment we are looking at is more like 1 - 2/game, so in essence, I wouldn't even worry about it. We will see less games where someone jumps from 80 to 160 in the last quarter, like we used to maybe 3 times a season, but those things are not factorable. It's very much a swings and roundabouts situations, because the weeks Gray didn't benefit from it, and others did, it in one sense robbed from his score.
Bottom line, I wouldn't let scoring changes like that influence your selections. They are not something we can plan on, or factor in, they are one of the random elements of the game, and the reduction of the factor in the "loading zone" actually reduces one of the elements of luck in SC, and rewards the players more consistent over the course of the game.

Rowsus · 3 Mar 2015

fodzilla said:

You're absolutely right, fod. Keep in mind, in those thrashings where the first quarter gets scaled up, we never see someones goal, and the play/mark/kick etc to get that goal, scaled up to 25 to 40 points worth. In the dying minutes of a close game, we have seen that more than once in the last couple of years.

Grant · 3 Mar 2015

What do you think Bellchambers needs to average to be considered worth selecting? And how much do you factor in his dpp status providing ruck backup?

NakinakiFC · 3 Mar 2015

Hi Rowsus. Have to say your work is incredible, have thoroughly enjoyed reading your analyses in 2015. My question to you is with regards to Nic Naitanui. From memory, NAB cup games haven't been too much of an indicator re rucks in terms of fantasy, with most coaches easing their prized no1 rucks into the season proper. Will you still consider Nic Nat if he only plays the 1 NAB game? I am aware that he is recovering from a minor back injury, I would have thought this kind of injury doesn't bode well for a ruckman a few weeks out from round 1! Cheers

Prochard · 3 Mar 2015

Hey rowsus,
I've been hearing a lot of coaches emphasiZe on the importance of selecting players who are in the top 10 for mid or the top 8 for fwd thus starting with a midfield like pendlebury ablett selwood rockliff or gray in the forward line due to the guarantee of them being in the top whatever in their respective roles. However, I really don't get this notion. As you have suggested in one of your earlier articles, supercoach is like a share trading game as opposed to be a competition of getting the best team possible. I do not see any reason why a team should pay overs and stack top players in different roles unless you can be benefitted from the captaincy scores. I would rather a midfield with a captain and a bunch of midfielders in the 15-20 range from the start of the season rather than a team full of ultra premiums which is only completed mid to late season.
My question is, if I were to start a midfield consisting of ablett jpk and 2-4 from Greene, greenwood, griffin, dangerfield, swallow, crouch, jack, Stevens what level of production should I be expecting and what averages will justify their selections while considering the fact that having a lmamy of them will limit my scoring later on in the season when most people have a full set of ultra premiums

happydays · 3 Mar 2015

Prochard said:

I agree.
The biggest concern I have with my team, is it so similar to so many others...

I feel like I need to back my instincts and turn a Shaw or BSmith into a 425k to 400k keeper; hope he puts in a 92ppg year and then use that 75k to upgrade a fortnight earlier...and give my team some distinction from the pack? And maybe do that twice?

Aware this is probably dangerous thinking...

Courtesans · 3 Mar 2015

Prochard said:

Hey rowsus,
I've been hearing a lot of coaches emphasiZe on the importance of selecting players who are in the top 10 for mid or the top 8 for fwd thus starting with a midfield like pendlebury ablett selwood rockliff or gray in the forward line due to the guarantee of them being in the top whatever in their respective roles. However, I really don't get this notion. As you have suggested in one of your earlier articles, supercoach is like a share trading game as opposed to be a competition of getting the best team possible. I do not see any reason why a team should pay overs and stack top players in different roles unless you can be benefitted from the captaincy scores. I would rather a midfield with a captain and a bunch of midfielders in the 15-20 range from the start of the season rather than a team full of ultra premiums which is only completed mid to late season.
My question is, if I were to start a midfield consisting of ablett jpk and 2-4 from Greene, greenwood, griffin, dangerfield, swallow, crouch, jack, Stevens what level of production should I be expecting and what averages will justify their selections while considering the fact that having a lmamy of them will limit my scoring later on in the season when most people have a full set of ultra premiums

Good question Prochard, look forward to what Rowsus may say. Cheers.

WandP · 3 Mar 2015

Rowsus said:

Thanks Rowsus! You remind me of my boss (in a good way) who always accuses me of "paralysis by analysis" when he sees me sweating for hours on a half percent increase in accuracy. Sometimes I get bogged down in detail and need someone to remind me of the bigger picture. Which you've clearly done. Cheers for giving me the bigger picture view! Even identifying a 5 point swing in average wouldn't be worth writing home about.

Nathan · 3 Mar 2015

Rowsus said:

Is 'absolutely right' a little bit less right than right?

Rowsus · 3 Mar 2015

Grant said:

Ok, let's see if we can take some sort of generic maths approach to this problem. Let's look at a PIT score/average for the F/R in your team. Keep in mind it is actually your F7/8 that is covering a Ruck outage, not your F/R.
Here is an equation that might help us.

DPP F/R "value" =

(Points scored by F/R + (F7 score x number of games missed by R1/R2 x Coverage factor) + F7 x number of games missed by F/R x coverage factor)/22

Looking at 9 popular starting Rucks last season (Jacobs, Goldy, Sandi, Cox, NicNat, Lobbe, Ryder, Minson, Mumford) they averaged exactly 20 games each for the season. So let's use that as the basis for the following calculations.

Assuming each of your R1 and R2 play 20 games for the season, and R1/R2 and F/R are playing Rnd 1:
The chances that the 4 games missed by R1/R2 occur in 4 separate rounds are 81.43%
The chances that the 4 games missed by R1/R2 occur in 3 separate rounds are 18.09%
The chances that the 4 games missed by R1/R2 occur in 2 separate rounds are 0.48%

If your R1/R2 miss their 4 games in 4 separate rounds, the chances your F/R is playing in those rounds are:
Playing all 4 rounds 64.76%
Playing only 3 rounds 32.38%
Playing only 2 rounds 2.86%

If your R1/R2 miss their 4 games in 3 separate rounds, the chances your F/R is playing in those rounds are:
Playing all 3 rounds 72.86%
Playing only 2 rounds 25.71%
Playing only 1 round 1.43%

If your R1/R2 miss their 4 games in only 2 separate rounds, the chances your F/R is playing in those rounds are:
Playing both 2 rounds 81.43%
Playing only 1 round 18.09%
Playing neither round 0.48%

So when we extend those numbers, we get the chances that you'll have cover for when your R1/R2 go missing.
Total coverage = (.8143 x .6476) + (.1809 x .7286) + (.0048 x .8143) = 66.31%
Coverage missing 1 game = (.8143 x .3238) + (.1809 x .2571) + (.0048 x .1809) = 31.10%
Coverage missing 2 games = (.8143 x .0286) + (.1809 x .0143) + (.0048 x .0048) = 2.59%
So when you start a F/R for coverage in your Rucks, assuming you always have a F7/8 available to replace your F/R in the Forward line, and your R1/R2 and F/R play the average of 20 out of 22 games, your coverage factor is only:

.6631 x 100% + .3110 x 50% + .0259 x 0% = 81.86%

Let's assume your coverage at F7/F8 is worth on average 70 points. It will be less than that early in the season, and hopefully more than that later in the season, when the Rucks are more likely to miss. We now have nearly all the figures we need to plug into the above equation.

(Points scored by F/R + (70 x 4 x .8186) + (70 x 2 x .8186))/22 = PIT70 average for F/R

=> (Points scored by F/R + 343.812)/22

Now we need to decide what we think is an acceptable PIT70 for our F6, as it is our F6 that F/R will be tipping out of our starting line up. If we assume our F6 will also play 20 games, and score at around 96, keeping in mind that the 6th best Fwd scorer this season might be up around 98 - 100/game, but we are not likely to end up with all of the top 6.
This gives us a F6 PIT70 average of:

(96 x 20 + 70 x 2)/22 = 2060/22 = 93.64/game.

For your F/R to achieve this he needs to score 2060 - 343.812 = 1716.188 points.
We have assumed he plays 20 games, so for your F/R to be a worthwhile pick, given the so many many ifs and buts we have used to arrive at this point, he needs to average:

1716.188/20 = 85.81/game.

This may seem low to some people, but remember, we are assuming that you have no other Ruck cover, and are trying to avoid trading around your R1/R2 missing. The points scored by your F7/8 when used as cover for a missing R1/R2 can in effect be attributed to your F/R, because without him, you'd miss those points all together.

So in general, and depending on exact circumstances, your F/R pick needs to average around 86/game to be considered a good pick, assuming no playing Rookie Rucks pop up in the meantime.

Interestingly enough, I think that is somewhere in the ballpark of what TBC will average this season.

Rowsus · 4 Mar 2015

NakinakiFC said:

Hi Naki, thanks for the kind words.

I will still consider NicNat if he only plays 1 NAB Cup game, but it would want to be full of positive signs. I would want to see high leaping, and good work below his knees.
I'm kind of torn with NicNat, He's really starting to become like that horse you back every time it runs, because he won that one good race, and keeps looking unlucky in the subsequent races. There comes a time when you just have to say "It's not bad luck, the horse just isn't as good as I thought it was, and that one win was just one out of the box". I'm not prepared to label NicNat "not as good as we thought he was" just yet, but he is teetering on the brink. He's played 108 games now, and yes he has a 20/114.2 in his form line, but he has nothing else over 100. Yes, he has been hampered by injury (hence his "bad luck" references), but let's look at it, he's played 76 games including finals since 2011, and only 3 of them are below 70% TOG. There's a nagging feeling in the back of my mind, that even when he is fit, he may never climb that 110+ mountain again.
Having said all that, I do believe when fit he will be no worse than a high 90's or low 100's SC player, and given he's priced at 91 this season, if we are ever to give him "just one more chance" this would seem to be the season to do it.
As of right now, he's not sitting in my team, but I haven't written him off yet. I just want to see something really positive before I bring him back in.

Rowsus · 4 Mar 2015

Prochard said:

happydays said:

If you can find $400-$425K Keepers, then do it, you'd be mad not to. Apart from that, don't worry about your team looking the same as everybody elses. There will always be differences, and trading will also create differences.

Courtesans said:

Hey Pro, happydays, Courtesans
it's a question that is actually more complicated than it looks, and actually requires an input from each individual coach, as to what they think each player will score, as well as what your intentions are if these players are all looking more like M8-15 type players as the season wears on. I could actually write pages on it, and still not cover all contingencies, so let's just try some bucket science on it instead.

The 8 players you mentioned have an average price of $521,812.50, so let's call them all $522,000 players.
The 6 players between Ablett and JPK have an average price of $653,950 so let's call them all $654,000 players.

So we are looking at comparing:
4 x $522,000 players = $2,088,000
3 x $654,000 players + $126,000 Rookie =$2,088,000

OR

3 x $522,000 players = $1,566,000
2 x $654,000 + $258,000 Rookie or $126,000 Rookie and $132,000 loose change for other areas.

Now lets make some assumptions/reasonable expectations.
You wouldn't/shouldn't consider the cheaper players unless you hoped they would average 105, and their shouldn't be any discussion about picking them, if you expect them to average 110 or better, they should be automatic selections. Se let's call them 105/game players. This is a happy coincidence, because in roughly round figures a 105/game player's price will tend towards the $522,000 area. So no loss, but no gain from them either.
We can't assume the more expensive players will score to last seasons average. The higher priced players struggle sometimes to reproduce last seasons figures. At $654,000 it assumes they scored at 121.7 last season. Let's just for argument's sake say they score at around 117 this season. Their price will tend towards $581,700 as the season goes on.

So we have:
$522,000 players, that scored at 97.1 last season, and will score at 105 this season.
$654,000 players, that scored at 121.7 last season, and will score at 117 this season.

We just have to assume that every player under consideration in this exercise plays all 22 games. It is pointless to factor in missed games and Rookies, as it evens out anyway.

So let's look at the 4 bargains against 3 top priced and a Rookie.

The 4 bargains will score 4 x 22 x 105 = 9,240 points for the season.
The 3 top pricers will score 3 x 22 x 117 = 7,722 points for the season.
This leaves the Rookie, plus his eventual upgrade, to score 1,518 points. That is only 69/game! Given we would be upgrading that Rookie somewhere between Rnds 6 and 15, to hopefully a 110/game player, it looks like a clear win to the higher priced players. The bargains haven't had any dollar growth, but by maintaining their price, the cost of upgrading one or two of them later on reduces to $40-$60k, as the prices of all players compress.
Let's assume the Rookie scores at 70, which is not ridiculous for a Mid Rookie, and we upgrade them in the byes at Round 12, to a 110 scoring player. This makes the top priced players 1 trade and $200k to produce:

3 x 22 x 117 + 11 x 70 + 11 x 110 = 9,702 points.

We had already established that the 4 bargain players would produce 9,240 points, if left un-upgraded. That leaves them 462 points behind. If we upgrade one/some to the 117 point players, we are now scoring 12/game higher, so we need to get 462 / 12 = 38.5 games out of the upgrades to finish in front of the top priced players. The top priced players prices are tending towards $581,700 and the bargain players are maintaining their $522,000 price, so each upgrade costs $59,700. To get the 39 games required out of the upgrades, we need to upgrade 3 of the bargains by round 9! ie one in Round 8, one in Round 9 and one in Round 10. The problem is, the top priced players probably won't have bottomed out by then, and you are using 3 trades.
So the bottom line comparison is:
3 Top priced players + a Rookie, using 1 trade and $200k = 9,702 points
4 Bargain priced players, using 3 trades and $179k (+ luck in Price) = 9,702 points.

Let's just take a quick look at the 3 bargain priced players against 2 top priced + Rookie + Loose change.

Bargain players: 3 x 22 x 105 = 6,930 points
Top priced players 2 x 22 x 117 = 5,148, or 1,782 short of the bargains.

The Rookie, if left untraded scores at 70/game, leaving the Loose change to score 1,782 - (70 x 22) = 242.
242 = 11/game. Loose change of $132k is expected to score at anywhere from 16 to 26/game in upgrades elsewhere. Another clear win to the top priced players. And don't forget, dollars hard to produce! In each scenario, the bargain players have 1 less dollar producing Rookie than the teams with the top priced players.

So where does this leave us, and doesn't this contradict the Share Trading scenario from last season?

Where it leaves us is, we have made some pretty general assumptions here. The bargain priced players will score at 105, the top priced players will score at 117. Of course they won't all work out that way. If you believe some of the bargains will score higher than 105, put a couple in, though I wouldn't suggest more than 2, unless you believe all the bargains you put in can go 110+. Similarly, you might be of the opinion that some of the top priced players will score more like 112, than 117. Then by all means, leave them out. Target them as upgrades on a fallen price. Depending on how your dollars are spreading around, I don't believe you should have more than one of these bargain type players in your Mid, unless, as previously stated, you are confident the ones you are including will go 110+. As to a player like Gray, I think if you have the funds you get him, if you don't, you hope he hits a bad score or two, and his price bottoms out. There are very few players in the Fwd line we can feel pretty confident they will score 100+, let alone give some hope to them going 110+. A compromised Fwd structure might see your F6 scoring around 4 - 6 below the better teams F6, a compromised Mid structure is more likely to leak more points. Take less risks there!

So, some of you may ask, doesn't this all contradict what I wrote last season? I think the answer is no, definitely not!
I have always maintained, that a bargain priced player needs to either achieve Keeper status, or produce points and dollars. The players we have just studied in the bargain bracket have really done neither of those. I believe you can possibly carry one 105/game Mid at M8, but that's it. If your M7 is scoring at 105, you're leaking too many points to the better teams.

By all means hunt a bargain, and if you think a $500k player can be a Mid Keeper, get them! (or $400k Def, $450k Fwd) Don't fall for the trap of pushing your expectations up. It is so easy to start thinking a player is 105, and then because you are in love with the price, you slowly talk him up to 110+. Nine out of ten times, this will end in tears! Set reasonable expectations, and only change them if you have good reason to, not because you are trying to talk yourself into having that player!!!

Rowsus · 4 Mar 2015

Nathan said:

Absolutely: without qualification or restriction. Completely.

Nahhh, I'd say it is about the same amount of "right" as right.

JMLP · 4 Mar 2015

Hi Rowsus, cheers for providing some awesome reading and insight in this thread!

Just curious of your thoughts on a couple players (apologies if you've already covered them):
Brodie Smith, Rory Sloane, Sam Docherty and Dustin Martin

And also curious to see your opinion on how much money people should start with in the bank (if any) for future trades once the season starts, I currently have about 100k in the bank and i'm wondering if it's worth potentially strengthening my team a touch more now or saving that to allow for more flexibility once the season gets going.

JT23 · 4 Mar 2015

Hi Rowsus

I know this question gets thrown around occasionally but was just looking for a bit of statistical thinking behind a structural decision.

In my midfield I have... Premium, Premium Premium, A, B, C, 200K Rookie, 200K Rookie. I am considering taking a slightly conservative approach with my starting midfielders and selecting players I personally believe have bigger roles to play in their own teams than their SC price suggest (i.e Van Berlo). SO I know some people will have issues with potentially too many 200K rookies - but that's ok if we can overlook that for the purposes of this discussion!

I have $970,000 and basically two options in my head.

A: Premium (Sloane or similar)
B: 200K rookie
C: Rookie (Jed Anderson / Heeney / whoever looks to be the best non-selected Rookie come round 1)

Or

A: Wells
B: another mid pricer (350K roughly)
C: 250K rookie (roughly)

Apart from the discussing about how many mid pricers you should have (esp in the midfield) and in the knowledge that I can select another KEEPER by taking option 1 (Sloane for example) how would I be best to evaluate the two methods in hypotheticals. I.E If I nail my mid pricers and they can average above 90 ppg (or even 100) vs saving a potential trade but having the lower base from which to fatten a cash cow.

Points v $$$ is an obvious comparison but Just wondering if we assign some points to players like these how would we evaluate this comparison.

Sorry for length of question - hope it is clear!

Opinion Questions For Rowsus

Rowsus

Statistician

Bobbie

Best and Fairest

Zim^zuM

WandP

150 Games Club

fodzilla

Rising Star Nominee

Rowsus

Statistician

Rowsus

Statistician

Grant

NakinakiFC

Prochard

happydays

Banned

Courtesans

WandP

150 Games Club

Nathan

Rowsus

Statistician

Rowsus

Statistician

Rowsus

Statistician

Rowsus

Statistician

JMLP

Rookie

JT23