RDM Poll time at 10 played!

[brian green]

The proposition that we "only" need five or six consecutive wins flies in the face of all accepted laws of probability.  If you have a horse that has run ten times and fallen at the last fence when in the lead five times, the odds of your horse suddenly jumping to victory faultlessly are hugely against and massively on more falls.  The form book does not lie, nor does the league table.

[Chico Hamilton III]

i just find it baffling that rdm, clarke and bond are all having an input and we still for the most part look like a sack of shit

I wonder what their relationship is like? A manager and 2 respected coaches with an expensively assembled squad in a relatively poor league and we're still chucking games away by conceding late goal after late goal. Has anyone noticed any real improvement? Who is making the tactical decisions here? Closing a game down, it's coachable but we're no better at it now than since the first game against Wednesday and I don't accept that it's something that requires more than 10 games to sort out. We might as well still have Lambert or any of the other shit managers we've had since.

And if I hear the word "gel" one more time......

It's unacceptable and we need to start winning in October or he can shit off. 

[kipeye]

Too soon. Not before the end of November. He has the makings of a team that need to build confidence and could start winning well. If nothing changes then before Christmas is about the right time but we will probably have chalked off this season by then anyway.

[Clampy]

The proposition that we "only" need five or six consecutive wins flies in the face of all accepted laws of probability.  If you have a horse that has run ten times and fallen at the last fence when in the lead five times, the odds of your horse suddenly jumping to victory faultlessly are hugely against and massively on more falls.  The form book does not lie, nor does the league table.

But the point is, it is possible to go on a winning run at some point. We just need to start doing it sooner rather than later. I suppose you could say there's still another 36 fences to jump.

[brian green]

Disagree Clampy I'm afraid.  There are not 36 fences to jump, there are 36 last fences to jump.  We have cast iron form for coping with all the other jumps but not the final ones.  Until we address and overcome whatever is causing our failing at the finish the odds remain very heavily stacked in favour of the future being the same as the past.

[Dribbler]

The proposition that we "only" need five or six consecutive wins flies in the face of all accepted laws of probability.  If you have a horse that has run ten times and fallen at the last fence when in the lead five times, the odds of your horse suddenly jumping to victory faultlessly are hugely against and massively on more falls.  The form book does not lie, nor does the league table.

Actually that's not how probability works, you do realise that even if you've rolled 5 sixes with a dice in the past 5 rolls the probability of rolling a 6 when you roll the next time is still 1 in 6?

[Boz]

All this talk of missing out on automatic promotion or even the playoffs is a bit melodramatic. I'd agree that I don't think we'll do it under RDM, but we're only 10pts and 7pts behind each of those targets so 4 or 5 consecutive wins will put us right in the mix. I think it will take a managerial change, but don't right off the season or promotion just yet.

We might be only 10 or 7 points off the top of the table targets, but we are a lot closer to the bottom three and on current form we look like remaining there unless a couple of wins are obtained. Even with these wins, we still need some of those above us to start dropping points, so my optimism of the play offs is not great at the moment, we need something more positive delivered by the team and manager.

[brian green]

We are not talking about random probability, we are talking about form.  Our form shows that we lose a disproportionately larger number of games at the death than would be statistically expected without cause or influence.  Until we identify and correct the cause of the aberration it is more likely to continue than to cease.

[Clampy]

For me it's not so much losing them at the death that's cost us, it's not having the game won before then, which in most games this season we should have done.

[Dribbler]

We are not talking about random probability, we are talking about form.  Our form shows that we lose a disproportionately larger number of games at the death than would be statistically expected without cause or influence.  Until we identify and correct the cause of the aberration it is more likely to continue than to cease.

No probability is exactly random, it depends on the number of possibilities vs causal factors you take into account. Plus we haven't lost a large number of games at the death, this season we've lost 2 games in the league, one (Bristol City) was lost way before the 'death', though the other (Sheff Wed) was lost in the 85th minute. Yes we've lost goals in the last 5 - 10 minutes of games far too much, ending up with a lot of draws rather than wins, which may make it feel like a loss, but lets be accurate about this.

The likelihood is, there is no single causal factor, missed chances not closing out games earlier on, defensive lapses, psychology, tactics and so on, another big factor is probably that of the psychology of other teams, who have seen this trend appearing and so it gives them the belief to push on for an equaliser or winner against us near the end.

[Chris Smith]

We are not talking about random probability, we are talking about form.  Our form shows that we lose a disproportionately larger number of games at the death than would be statistically expected without cause or influence.  Until we identify and correct the cause of the aberration it is more likely to continue than to cease.

Draw rather than lose, surely.

[Villa in Denmark]

We are not talking about random probability, we are talking about form.  Our form shows that we lose a disproportionately larger number of games at the death than would be statistically expected without cause or influence.  Until we identify and correct the cause of the aberration it is more likely to continue than to cease.

No probability is exactly random, it depends on the number of possibilities vs causal factors you take into account. Plus we haven't lost a large number of games at the death, this season we've lost 2 games in the league, one (Bristol City) was lost way before the 'death', though the other (Sheff Wed) was lost in the 85th minute. Yes we've lost goals in the last 5 - 10 minutes of games far too much, ending up with a lot of draws rather than wins, which may make it feel like a loss, but lets be accurate about this.

The likelihood is, there is no single causal factor, missed chances not closing out games earlier on, defensive lapses, psychology, tactics and so on, another big factor is probably that of the psychology of other teams, who have seen this trend appearing and so it gives them the belief to push on for an equaliser or winner against us near the end.

Maths 101.

Probability isn't random.

This is a standard distribution of results.



the x axis is the number of games where points are thrown away in the last 10 minutes and the y axis is the number of teams throwing those points away, a random event would expect a result somewhere close to the centre line. 

Ours is so far to the right, i.e. not many teams throwing loads of games, it's almost off the scale and if you can be bothered pissing about with processing all the data and all the standard deviation calculations I reckon you'd come up with a >99% statistical certainty of a causal effect.

[Jimbo]

I think we probably need to accept that the previous owner left us in such an unholy mess that this rebuilding job will take at least two seasons rather than one.

In other words, forget promotion and playoffs. Forget them completely. Concentrate on each game as it comes and see where it sets us up for next season. Try not to get too frustrated when the likes of Newcastle and Norwich are getting results. We were in much, much worse shape than them and it's clearly not an overnight job putting it right.

[Holte L2]

And yet nearly all of us were happy with that starting 11.

I wasn't mate. Midfield was/is far too lightweight.

[brian green]

If you throw dice the laws of probability dictate that each number from one to six will arise one sixth of the time provided a large enough number of chances form the sample.

If you go dog racing and there are six dogs each in a trap from one to six, over a sufficiently long sample traps three and four never win as many races as trap one or trap six.  That is because extraneous factors such as railing and wide running come into play.  Different greyhounds lead with a different leg making them wide runners or railers.  Extrapolating a dog's racing habits enables you to take such factors into account.  That is form.  The fall of dice is not affected by previous events unless of course the dice are loaded and that is an entirely different debate.