Do We Survive?

I hope we win this one

 

I also hope this.

 

The three outcomes (finishing above Burnley, finishing above Norwich and finishing above another side) are heavily correlated. They are not three coin flips where the chances of 3 Heads is 0.5x0.5x0.5 (12.5%). Instead if we finish above one of the other sides the chances are very high that we will have finished above Burnley and Norwich.

I think we are agreed that there is a 40% chance we finish above one of Leeds/Everton/Brentford/Newcastle. If we do that it’s because we’ve had an upturn in form and got say 33-35 points +. If that happens we will almost certainly finish above Burnley and Norwich unless they have a similar dramatic upturn in form. I would put the probability of one of either of those sides having a dramatic upturn in form that outstrips our dramatic upturn in form at maybe 1/4 which would put our probability of staying up is 3/4 x 0.4 which is 0.3 or 30% not 8%.

I don’t think there is much need to get too mathematical about this though. I think any football fan intuitively knows that our chances of staying up in this situation (a third of the season left, four points adrift of numerous teams who are struggling) seems decent if unfavourable, much more like 1 in 3 than 1 in 12. And that’s before you factor in that we play Burnley, Everton, Leeds and Brentford, all at home.

I still think we go down. But staying up isn’t that improbable yet.

 

I get your point. I've treated the outcome as being 3 separate events. If you need x+y+z to happen, the probability is x% * y% * z%. ie 3 heads when you toss a coin is 50% * 50% * 50% = 12.5%. . They are separate events to the entent that each of those clubs may have great form, average form or poor form, however the common event to them all is our form. If we went on a truly great run then to an extent their form becomes less relevant. Likewise if we maintain our form of the last 12 games or so, their form is also less relevant. Other than maybe Norwich, who of course thrashed us, I'd argue we are less than 50/50 to finish above each of the others, and we are unlikely to go on such a great run that others' form becomes irrelevant. Who know but my gut feel says finishing above 3 other teams (Norwich, Burnley & one other) based on what we know today is about 10%. Beat CP and it's up to maybe 25%, lose and it's down....

 

As last Saturday showed we are certainly capable of staying up, beating Palace on Wednesday would go a long way, we have tossed too many points away we need to button down the hatches and bring home the bacon.
Three points on Wednesday and another one on Saturday wouldn’t go amiss!

 

This all hinges on the likelihood of us having a great upturn in form. That is baked into my %s and I think it's pretty unlikely. It only needs one of Burnley & Norwich, rather than us, to have that great upturn and we are effectively dead. I think the maths helps show how much less likely it is that we will finish above 3 teams as opposed to finishing above one.
EDIT - used the 3 coins example before seeing you'd used the same !

 

You don’t appear to have factored in the impact on performances that forum members engaging in a collective bout of positive thinking might have. Back to the drawing board.

 

I believe the constant repetition of ‘the future’s bright, the future’s yellow’ brings with it an additional 1.7% chance of staying up.

 

8% chance of staying up suggests we'll finish bottom & the flaw is to multiply 3 inter-related but individual probabilities together. About as valid as taking the average of the same 3 individual probabilities, which would give a result of 43%.
If we start from the premise that we are not likely to finish above 17th & that there are 5 other sides in the mix, we could calculate our chances of finishing above any of them as:
Norwich: 50%
Burnley: 40%
Newcastle: 35%
Leeds: 30%
Brentford: 25%

Average = 36%, very close to the 4/11 Diamond (I think) had the bookies' odds at for us going down.

The declining scales of probability I have chosen reflect that we cannot be equally likely to finish above all those teams, especially considering they will be playing each other in a lot of games.

Edit: forgot Everton (continuing hangover from the Cup Final I suppose).
If we put them in at 25%, then the likelihood of surviving comes down to 34%; put them in at 30%, it comes down to 35%. All still in line with the bookies.

 

Why aren't Everton in the mix?

 

Careful, mate; there's only so many mixed metaphors left in the world, you know!

 

See my prompt edit.
I keep hoping they'll disappear....

 

You can get about 30% return with the bookies on us going down. Sounds like it’s time to remortgage the TuT residence!

 

Shouldn’t adding Everton to the mix increase our chances of saying up? If so then your methodology seems a little flawed. But intuitively that’s what I have it as too (although I think Newcastle are looking stronger than Everton).

 

I think your odds for the last 3 are too optimistic unless you are factoring in that you believe we will have a serious uplift in form.
Norwich: 50%
Burnley: 40%
Newcastle: 15%
Leeds: 15%
Brentford: 20%
It still doesn't feel right to say we have a 50/50 chance of finishing above one of the latter 3 but maybe if we add Everton in, so let's use that and therefore the average is 31% for argument's sake.
But your average is of course meaningless.
If we had to finish above only one, the average of the 5 is still 31%. If we had to finish above all 5, the average is still 31%. Do you think both outcomes are equally likely ? If we add the next 3 clubs in the average will drop further, do you think we would be less likely to stay up if we looked at 8 instead of 5 clubs as the average would be lower ?
Realistically we will need to finish above Norwich and Burnley and one of the others and you have quoted N&B each at 50% or lower.

The only way to get an agreed position on % likelihood would be to fix one of the variables eg;
1/assign a % likelihood to Watford getting each number of possible points, then
2/ assign a likelihood of finishing above each other team under that scenario.
3/ repeat for all other possible number of points, then
4/ multiply out the whole lot.
I'm not proposing we do this

 

That is similar to what the bookies (4/11) or fivethirtyeight (29%) do although as I pointed out earlier the latter puts too much weight on their perception of a team’s quality which is why their model underrates Brentford’s chances of going down but is broadly correct.

 

Bookies odds of course always favour the bookies - they aren't their expected outcomes ! Supports that the real odds are far lower.

 

There’s obviously a house cut but(and I don’t bet so don’t know this kind of thing) there is almost certainly some expected probability underlying it that wouldn’t be too hard to figure out if you looked at other odds.

 

@Otter would know what margin they would add into the odds but if we are saying that finishing above Norwich is 50%, Burnley 40%, and another side maybe 25-30%, I'd be amazed if the probability of all 3 happening is above 20%.

 

He was pretty ineffective against us so let's hope he has plenty more games like that, as well as that their 3-0 at Brighton turns out to be a freak result like their 4-0 at Wolves last season.

 

If we don't have a serious uplift in form this thread is a waste of time!

 

...and what do you think is the % of that happening ?
Burnley have 4 games in hand on Brentford and are 7 points behind. They meet on 12th March. I think if Brentford lose that and don't pick up points in their next few easy games, the 3 promoted sides will go back down.

 

It depends on the sport, the number of 'competitors' and the number of outcomes. Most sports' sum of odds favour the bookies by around 110% to 120% with a single outcome, however with a two-way outcome it can be a bit less, e.g. 1.5 goals, more=10/11, less=5/6 equates to about 107%.

For the purposes of the the odds for relegation, there are 3 "winners", it becomes a lot more complicated and a greater risk to the bookie. Most money is made on favourites and other short priced competitor but also carries the greatest risk to the bookies if they don't manage their liabilities properly. When a 100/1 horse wins a race, it's good for the odd individual who put a bet on it, but very very good for the bookie as they would have cleaned up on the favourites.

 

How does the gambling industry calculate probability of staying up accurately? Does it factor in form, talent, injuries, weather, fixture congestion and the million other variables? Or does it ultimately come down to a best guess?

 

At this stage of the season the number one driver would be liabilities. If they potentially have to pay out a large sum on a team, they will advertise unfavourable or non-competitive odds on that team and offer tempting odds on an outsider in the hope of clawing some money back.

 

bump

 

Also bump

 

After the last 3 home matches, it's an emphatic no. The club and players don't give a damn.

 

Not quite sure I agree with the players not caring. I just think there's a woeful lack of quality, and the majority of the squad simply aren't good enough.

 

Gloat as much as you like. I knew you and your other aliases would be in clover tonight. Enjoy it. You'll all be suicidal again if we win at Old Trafford. At least you were quiet for a few days after the Villa win.

 

Of course players care

But their heads have gone, and taken any belief and confidence with it

 

Even by the standards of our COG fanbase I’m astonished at how many seemingly decided tonight that we are now probably going down

These clowns were as good as relegated a month ago

 

You’d have thought the Norwich game might have made them twig…