NRL Fantasy 2021 Part 4 - Praying that nobody gets hurt

RandomSil wrote:

Fractionally underpriced due to five games below 60 minutes with four of those under 40. Averaged almost 54 points while being priced at 46. Had a fair few tries though. Good lowcut keeper potentially

RandomSil wrote:

Fractionally underpriced due to five games below 60 minutes with four of those under 40. Averaged almost 54 points while being priced at 46. Had a fair few tries though. Good lowcut keeper potentially

L-Jimmy wrote:

MattZN is right about injury.

But even so - this is what folks mean by cherry-picking stats. People get injured and have short games due to HIA - you shouldn't dis-include those points from the sample as they are likely to happen again. Sure, if they are played back in to manage an injury over a period of many weeks then make adjustment for that.

Sure, a player may have fewer injuries/HIAs this year. They might have more too. Using as big a sample size as possible will (all else equal) move you closer to a representative sample.

my tv broke wrote:

Locked in for left centre Round 1.

my tv broke wrote:Speaking of RFM and the Dogs, I reckon there is a chance Waddell locks down a starting spot, and I'd probably run with him if so. Interview on the radio the other day he said he was looking forward to getting the ball in his hands more this year and doing more in attack, which was something he found hard to do coming off the bench last year.

Not sure how they would fit but there has been rumours JJ would move to lock..

One to keep an eye on for me.

mattnz wrote:
Removing the outliers is a way to normalise everyone for comparison. Getting injured 2 minutes into a match shouldn't penalise more than getting injured 78 minutes into a match, or getting injured at training.
Having said that, a guy like Turbo who has had 5 separate injuries in the past 2 seasons that have kept him out for weeks should have a bearing on selection.

Also note that I am saying RFM is injured and going to miss round 1 this season

my tv broke wrote:

Well, he played 80 minutes twice.

New coach, positions are up for grabs. Backrowers could easily be JJ and Waddell too

mattnz wrote:
Elliott was one of their best and was playing 80 at lock before getting injured, hard to see him being moved.

my tv broke wrote:
- patience, don't need to chase every cow or big score.

easytiger wrote:Just on Luke Brooks;

I'm a little mystified why people think he will get a significant boost in his Kick Metres?
Over the past 7 years in 70+ minute games (142 of them) he's averaged 233KM.

His average over the past 3 years playing next to Benji is 256KM
His average over the past 3 years when playing without Benji is 265KM (essentially barely any difference).

His average prior to Benji coming to the club was 214KM.

He's kicked over 400 metres in a game on 6 occasions across 142 games.
He's kicked over 300 metres in under 30% of his games (to put that into context, Milford's quiet year last year where his KM's were massively down, still had him kicking over 300KM in just over 30% of his games). There's no real point comparing him to the big KM halves, because it's just not even close to being comparable.

Benji's average playing next to Brooks was 195KM.
Adam Doueihi averaged 235KM from 5 games playing in the halves - which suggests he is at the very least competent enough to keep doing in-play kicking - so there's no real reason that I can see why Luke Brooks should suddenly dominate the kicking.
I'd also mention that the pre-season "It's Luke Brooks' team to run" has become an annual press release for a number of years now...

Don't get me wrong, I think Brooks should have upside, a little from slightly reduced minutes last year (still averaged 70 minutes per game), but it seems to me that any upside is most likely to be from boosting other attacking stats back to their normal levels? - which kind've makes it a hunch pick

L-Jimmy wrote:

Outlier removal is risky and must be done parsimoniously; there is a massive opportunity to introduce analyst confirmation bias. Best practice is to only do it for the most extreme - 1min HIAs maybe.

But even those aren't low-probability events really - we often see a couple each round. Removing them is effectively forecasting their absence, and poor practice.

tl;dr doing good stats is hard and relying on averages is sensible.

Mulvy wrote:

Yeah, have to be sensible about it. I use 60 mins for backs because occasionally they will get subbed off towards the end of a game. Excluding 10 min injury games is also wise. But someone like SST who played a few one stint games for 20 odd minutes when the Roosters were up by miles, you need to leave those in because they are likely to continue.