GT49er
Well-known member
See post #10. Original discussion about point per game - other variations (like 40 points, 1.05 per game, etc.) came later.
Got something to say or just want fewer pesky ads? Join us... 😊
[Albion] To Avoid Potter Bashing - the staying up tracker - season 2019-2020 Game 38 UPDATE
- Thread starter Giraffe
- Start date
Nixonator
Well-known member
We should probably sign some players if we expect either of those trajectories.
Bob!
Coffee Buyer
- Jul 5, 2003
- 11,119
See post #1 - Original aim of thread to get to 40 points, and then post #36 graph showing how to get to 40 points.
The Brighton Bear
Come on Kylie, get a grip
What about an average boot size per team per match line?
(Obviously compared with last season)
(Obviously compared with last season)
GT49er
Well-known member
Post #1 was all about Giraffe's excellent tracker; no criticism from me about that, and no debate either. The debate was about the 'alternative' tracker that Thimble Keegan was advocating, which was one point a game.
So, to simplify, the aim of Giraffe's checker (not based on any average number of points per game) was to get 40 points.
Thimble Keegan's model, on the other hand, the one which I and several others were discussing, was based one point per game (adding up to 38 points); many of us pointed out that it was a crap idea.
Does NSC have to crap on every single thing that's good about this place?
FatSuperman
Well-known member
- Feb 25, 2016
- 2,830
But that is exactly how we know it is good. There have been precious few “we won’t get x points from y games” responses so far, but it’s early days
Blue line for actual on the very top please!
FatSuperman
Well-known member
- Feb 25, 2016
- 2,830
We definitely need one showing the various super-computer predictions! Like this travesty;
https://talksport.com/football/5578...per-computer-predicts-table-after-five-games/
- Aug 8, 2005
- 26,544
- Thread starter
- #71
Which gives us one point from five games. Eek.
um bongo molongo
Well-known member
The logic of the Giraffe tracker I think worked for the Hughton era, where we seemed to “know our place”. We generally (last few months notwithstanding) beat sides in the bottom 10 (or sides “at our levels”) at home, got the odd result against the top sides at home and lost to pretty much everyone away. I have a feeling the Potter era will be far more unpredictable. Maybe we need a “Potter rollercoaster” line which allocates points at random by game to get to 40 at the end of the season?
FatSuperman
Well-known member
- Feb 25, 2016
- 2,830
I’ll support this extra line. You now need two more sponsors to push it through.
Caveat is the random points number must be between -1 and 4
Quick work [MENTION=3734]Giraffe[/MENTION].
I see you've got us down for Hughton-esque runsin Nov/Dec and Mar/Apr. Hopefully Potter pulls off some surprise results.
He hasn't got us down for anything other than we beat the bottom 9 teams at home every time and draw with them away. While losing every game to the top 10 teams. This is hardly spectacular analysis.
London Irish
Well-known member
7 points from the first 3 games. No pressure then
sussex_guy2k2
Well-known member
- Jun 6, 2014
- 3,733
Asleep?
MattBackHome
Well-known member
- Jul 7, 2003
- 11,723
I actually tweaked Giraffe's tracker last season, incorporating the H/A points ratios of opponents to tweak exactly which games required wins and draws. The end point of the line was obviously the same but the shape was marginally different. Examples of results that my model predicted were the away wins at Palace and Huddersfield. If we're judging the models by average variance from actual then mine was marginally better*, though I did find us 10 points above expectation at one point (I think G's tracker had us at +7?) so the main tracker never deviated as much as mine did.
I'll do the same thing this year and add it in here seeing as everyone and his dog has their own model these days, no matter how shit they are.
*aftertimer alert
I'll do the same thing this year and add it in here seeing as everyone and his dog has their own model these days, no matter how shit they are.
*aftertimer alert
- Aug 8, 2005
- 26,544
- Thread starter
- #78
After careful consideration and listening to the opinions of no one else I concur. Here you go.
Attachments
Hark at the newboy getting all ahead of himself.
You've only been on here for 16 years - you need to do your time before you start chipping in with stuff like this.
But Giraffe's tracker is utterly pointless in that it doesn't matter how many points we have after 1,5,10,15 games. It's what we have after 38 that is the ONLY sensible measure. If we have 38, then the chances are we will stay up.
Simple.