The side is still in a magnificent position

[Albert]

Losses always hurt, especially at the top end of the table, and especially when an opponent has had it over you for a while. Ultimately though one match doesn't make a season, nor does two, nor half a dozen or even a bakers dozen. No, a season at this level is made up of 46 matches, two against each opponent, once home and once away. With 25 games played, we stand in the best position we have stood in for many years, and one that even with a poor performance from here on out will more than likely land us a result that most would have not only taken at the stand of the season, but would have hailed as a fine achievement.

As things stand we are still 8 points clear of 6th, have 10 points in the last 6 games (which at about 1.67 points per game is still playoff form) and have now faced the league leaders, and likely Champions home and away. Ultimately, there is a long way to go, and anything could happen, but if you offered me this position, even with the result last night, I'd have taken it, running all the way home laughing. But let's consider the facts in depth for a moment.

We have struggled against the top sides in the division, that is something that is a concern, but something that can be corrected. As it stands we have played 6 matches against the current top 6 and have amassed but a single point, away to Ipswich in a 4-4 draw. With Burnley away, QPR at home, Forest at home and Ipswich away still to come, and a high likelihood of the playoffs, this could certainly be where the wheels fall of this season.

The numbers do tell another very positive story as well though. Against the rest of the league we've only dropped 14 points from 54, that is, we've picking up points at a rate of 2.22 against the teams outside the top 6. What's more, whilst we sit decidedly mid table in the home table we stand atop the away table proudly in a vast turnaround from last season. We don't just stand atop it for this season though, our current away record stands proudly as one of the best in recent times. There's still 10 away games to go, and if we could pick up 20 points from those, that is at the current rate of scoring, it would be a magnificent achievement.

As for more general analysis. Using an updated form of the analysis performed on the table, we get around a 96% chance of the playoffs, taking yesterday's result into account. Even losing the next 3 we'd still have around a 79% chance. Below is a plot of probability against placing using the next form of the Championship table:

"lVsKvy9" alt="lVsKvy9">

As before, this is a plot of probability against league position, with green representing automatic promotion, blue the playoffs and red relegation.

There are two major changes to the previous analysis, both to do with how points per game is converted to league placing. The first major change is how the current shape of the table was used to predict the final table. Using ten years of data the progression of the table was modelled as a function of games remaining and position, and as expected the general trend was that the table would "flatten out", that is, top teams performances dipped, bottom teams got better, and the teams chasing the playoffs performed better as well. This analysis isn't perfect, but it gives a decent guide to how the table will most likely evolve over the course of the season. It should technically offer some insight into how the final table might look as early as 12 games. Be careful to note that this analysis is independent of which teams occupy which positions, that is, it's attempting to predict the approximate points totals of positions, not the points totals of the teams in those positions at the time. Just as a demonstration of what kind of predictions this makes, below is the last ten seasons final table (the left of each pair and grey) and the prediction using the table at this point in the season (right and green). Carefully note that whilst there are some that end up particularly wayward (first place in 08/09 and 02/03, 6th in 05/06 are particularly off), it generally is a decent guide. Below are the ten seasons worth:

"0d94HY5" alt="0d94HY5">

As for a better analysis of how accurate this is, overall the standard deviation the error is 3.64. This means that you'd expect the values to be within ±3.55 points in 68.2% of cases. For example, if 6th was expected to be 70 points one season, it would be expected with 68.2% certainty that the true final points would be between 73.55 and 66.45. Of course you can't have decimal points. For more practical certainties see below:

±1 point: 22.2%

±2 points: 42.7%

±3 points: 60.2%

±4 points: 74.1%

±6 points: 90.1%

±9 points: 98.9%

±12 points: 99.93%

Which ultimately isn't too bad. I might do a full analysis later to determine how this accuracy chances the closer to the end of the season it is so these errors can be used to compare later. In any case, for those interested, the current predicted final table is:

"v46aDg2" alt="v46aDg2">

In any case, the second change to the analysis used for prediction the chance of final placing is using a more accurate model for table shape. It's quite a simple change, basically just drawing straight lines between the positions rather than trying to force some smooth function of in (which can cause issues). Below is a comparison of the two methods, the blue being the new model, the dotted red being the old polynomial one, and the green being the predicted league finishes as above.

"KXrsw8t" alt="KXrsw8t">

As for what that plot way back when... uhh, this one:

"lVsKvy9" alt="lVsKvy9">

...actually meant; basically our position is pretty much the same as it was. Again, one loss doesn't really change anything at the moment, we still have 21 games to play, we've in a very strong position, and ultimately it's unlikely that we'll miss out on the playoffs. We'd need at most about 32 points to guarantee the playoffs as thing stand, about 1.52 points per game, which a team in our current position should be more than capable of getting. Even if we couldn't though, currently it's predicted that there will be quite a low playoff score this season, around 67 points, which would be a very low score for the playoffs, but after last season we know that low scores can and do happen. If it were only 67 the side would be a mere 23 points off, and from 21 games we'd need only to not drop into relegation form. This all said of course the aim should be to push for second place or beyond, although Leicester have more than likely put themselves into a position to seal top spot this season.

In terms of probabilities as predicted by the analysis (position or higher):

1st place: 8.35%

2nd place: 29.6%

6th place: 95.7%

21st place: 99.999%

As a side note on Leicester, they are given a 82.4% chance of finishing top now and 96.6% chance promotion and 99.999% chance of playoffs. 7 points clear though, I guess it didn't take this kind of thing to tell anyone that. Whilst I would say it would take a disaster for us to not finish in the playoffs, if they miss out on automatic it would take an collapse of amazing proportions for them to miss out on a spot in the Premier League, we'll see though. Good luck to them I guess.

So yeah, we've most definitely in a magnificent position, and I guess the one thing to take away from all of that, if anyone's actually bothered to read it, is that this is the kind of times we should sit back and enjoy. You never know what'll happen this season, you never know what'll happen next, but right now we're in the best position we've been in the best part of a decade. Soak it in, enjoy it, and get behind the side, whether this is our season or not, these are the times to enjoy. The side is playing some good stuff, we have some young talented players out there, and even if some matches don't go our way, we've had a great season so far.

[maydrakin]

Excellent post.

We have 21 games to go and all those around us all still have Leicester to play.

It is worth adding that, bar QPR, our defeats to the top sides all occurred pre-McClaren, so it's not yet a time to throw the baby out with bath water!

[Ramarena]

If we beat most of the teams below us, we will be fine. 

[Boycie]

Let's see if the ownership group are willing to stick or twist. If they done want to invest in promotion with the team in this position, then they never will.

The promised land is within touching distance, and if we don't get there, I feel the team will be broken up and sold.

[TigerTedd]

Let's see if the ownership group are willing to stick or twist. If they done want to invest in promotion with the team in this position, then they never will.

The promised land is within touching distance, and if we don't get there, I feel the team will be broken up and sold.

Not necessarily. McClaren was bought in for three years with a plan. Our resurgence has taken wen him by surprise. Similar to when Dopey the Dwarf took over, he had a 3 year plan, and promotion in year one took us all a bit by surprised.

Of course I'd welcome promotion whenever it came, but if it didn't come this season, that wouldn't mean that the wheels have fallen off the plan.

That doesn't mean the team won't be broken up, but if it is, it's because it's part of Steve's plan to build his own team. Although of like to think that by now he's got the idea that he's been lucky enough to inherit a very good team here an he doesn't need to do a great deal in terms of totally rebuilding a new one.

[Gritstone Ram]

Wow good post albert. Far too long for me to read. Good graphs anorl.

[RamNut]

So the last analysis was ****** then.

[Albert]

So the last analysis was ****** then.

No, comes to roughly the same conclusions. Just working on the finer details of the analysis. For comparison the revision of our automatic promotion chances is from 31.6% to 29.6% with the new analysis.

[Day]

"3kLn7sL" alt="3kLn7sL">

[RamNut]

No, comes to roughly the same conclusions. Just working on the finer details of the analysis. For comparison the revision of our automatic promotion chances is from 31.6% to 29.6% with the new analysis.

Previously it was showing 2nd as the most likely position which was odd .......now its 4th.

[Albert]

Previously it was showing 2nd as the most likely position which was odd .......now its 4th.

This plot:

"iCtNM31" alt="iCtNM31">

It appeared on the plot as position 1.56, but it should have been 2.56, I had messed up the graphic somewhat (part of the motivation for fixing a few things). The underlying analysis had picked 2.56->3 as the most likely position from that data. The plot should have looked more like:

"2XLKvgW" alt="2XLKvgW">

The key thing to note is that the added resolution to the analysis, particularly about the top end of the table, does have a few striking differences. The most notable are in the boundaries between the playoffs and the rest of the table, but it has also changed the shape of the top two, this being important here. Below is the above plot using the new analysis:

"8XzZQoA" alt="8XzZQoA">

The old analysis gave the most likely finish as 2.56, whilst the newer version gives it as 3.14 (~π).

The differences here are issues of points per game to league position, which is a hard question at the best of time, but hopefully the improvements help.

[PistoldPete2]

Albert I did reply to your last similar statto anaylsis saying I thought you were overestimating our chances, and I still do.

First thing to say after last night is we have done amazingly well to be in an excellent chance of a play off place.

But...

 

our record against the top teams is still very poor. teams like Wigan , Blackburn, Bolton and Brighton are all getting back to form, and unfortunately we have to play them all in the coming games.

I can see a very sticky patch in Jan/Feb and we could already be 6 points off an automatic place by 5pm today . Really think automatic promotion is  long shot now.  

 

 

[RamNut]

I also think we've lost the plot a bit.

[-JW-]

Can someone summarise Albert's excellent post in two sentences please??

[Albert]

Albert I did reply to your last similar statto anaylsis saying I thought you were overestimating our chances, and I still do.

First thing to say after last night is we have done amazingly well to be in an excellent chance of a play off place.

But...

 

our record against the top teams is still very poor. teams like Wigan , Blackburn, Bolton and Brighton are all getting back to form, and unfortunately we have to play them all in the coming games.

I can see a very sticky patch in Jan/Feb and we could already be 6 points off an automatic place by 5pm today . Really think automatic promotion is  long shot now.

As mentioned in the post, we have been poor against the teams currently in the top 6, but our form outside the top 6 has been nothing short of sensational. Whilst we might have a hard run coming up, we ultimately have played the two hardest matches of the season, and only have 4 more against those opponents currently in the top 6. Having only lost to 6 opponents so far this season, and only having 4 meetings with teams that have beaten us, I think you're overestimating the effect that the teams we have struggled against will have on our season overall even if they did go ahead and get the double over us. Even if we lost the next 5 games on the trot (that is, we lost to Burnley, QPR, Forest and Reading, as well as throwing Brighton, who we beat away, in for good measure, we'd still likely be just outside the top 6, in 7th or 8th, we still around a 50-50 chance of making it the way the table is shaping up.

This all said, the real worry becomes the playoffs if we do actually make it. Having checked over the last 10 years worth of playoffs it becomes painfully clear that the records of teams against eachother that season seems to make a massive difference. If you consider the matches between them over the season as a 4 team group stage of sorts, the top team of that group won the playoffs in 5 of those 10 seasons, with second picking up 3. That's said, I doubt people need to hear that to know that having a poor record against the top teams is hardly good news for a side in the playoffs. This all said, we did lose most of those matches before Steve came in, so who knows.

All things considered though I wouldn't read too much into the Leicester game, or the Wigan game. If we did in fact go on to lose the next 5, yeah... that's probably panic stations, but a couple of losses doesn't change the achievements of the first half of the season, and even a poor performance from here on out is likely to land us a playoff place. If we do match the kind of form that the side did produce in the first half of the season though, we are serious contenders, particularly with our amazing away record and apparent strength against teams lower down the table. We'll see though.

As a side note, I wouldn't exactly rate ~30% as a likely event though. If we were to have this position at this time of year for the next 3 seasons, you'd only expect us to go up in one of those still.

[PistoldPete2]

As mentioned in the post, we have been poor against the teams currently in the top 6, but our form outside the top 6 has been nothing short of sensational. Whilst we might have a hard run coming up, we ultimately have played the two hardest matches of the season, and only have 4 more against those opponents currently in the top 6. Having only lost to 6 opponents so far this season, and only having 4 meetings with teams that have beaten us, I think you're overestimating the effect that the teams we have struggled against will have on our season overall even if they did go ahead and get the double over us. Even if we lost the next 5 games on the trot (that is, we lost to Burnley, QPR, Forest and Reading, as well as throwing Brighton, who we beat away, in for good measure, we'd still likely be just outside the top 6, in 7th or 8th, we still around a 50-50 chance of making it the way the table is shaping up.

This all said, the real worry becomes the playoffs if we do actually make it. Having checked over the last 10 years worth of playoffs it becomes painfully clear that the records of teams against eachother that season seems to make a massive difference. If you consider the matches between them over the season as a 4 team group stage of sorts, the top team of that group won the playoffs in 5 of those 10 seasons, with second picking up 3. That's said, I doubt people need to hear that to know that having a poor record against the top teams is hardly good news for a side in the playoffs. This all said, we did lose most of those matches before Steve came in, so who knows.

All things considered though I wouldn't read too much into the Leicester game, or the Wigan game. If we did in fact go on to lose the next 5, yeah... that's probably panic stations, but a couple of losses doesn't change the achievements of the first half of the season, and even a poor performance from here on out is likely to land us a playoff place. If we do match the kind of form that the side did produce in the first half of the season though, we are serious contenders, particularly with our amazing away record and apparent strength against teams lower down the table. We'll see though.

As a side note, I wouldn't exactly rate ~30% as a likely event though. If we were to have this position at this time of year for the next 3 seasons, you'd only expect us to go up in one of those still.

 

Yeah but I think our chances of automatic is only about 10%.

I am not just concerned about Fester, Burnley, QPR, Reading  and dare I say it Forest doing the double over us I worry about us losing to Blackburn, Leeds, Brighton etc, just as we did to Wigan  . If we do , then with 22 games to go. we need 30 points say 10 wins for playoffs, or 9 wins and 3 draws- I think we will get that but only just and automatic is out of reach I think.  Theres no wiggle room to allow for banana skins against the lesser teams.

 

 

[randombc]

Brighton, Blackburn and Yeovil still to go this month,

 

If we get 5 or more points from them 3 everything will be rosy again!

[Albert]

Yeah but I think our chances of automatic is only about 10%.

I am not just concerned about Fester, Burnley, QPR, Reading  and dare I say it Forest doing the double over us I worry about us losing to Blackburn, Leeds, Brighton etc, just as we did to Wigan  . If we do , then with 22 games to go. we need 30 points say 10 wins for playoffs, or 9 wins and 3 draws- I think we will get that but only just and automatic is out of reach I think.  Theres no wiggle room to allow for banana skins against the lesser teams.

The idea of the analysis is to try and estimate the probability of various finishing positions. At the current time the way this works is by using known deviations from the the team's points per game from the season so far, and use that to determine the likely points by the end of the season. Currently that calibration was performed for the entire table, and there is an argument to redo it by position, but the key is that the current calibration has already shown that these predictions are at the very least in the ball park.

In terms of the playoffs, the current prediction is more like 68-70 points, although 75 should always be the aim. At this point though we only need midtable form to break that 75 point boundary (so the playoffs upper requirement).

Speaking just about automatic promotion, at this point we stand in 4th, and as one of the three real possibilities. Even a heuristic analysis would state around a 1 in 3 to 1 in 5 chance of us making that spot. This all said though, I think there are still two key things to be fixed up in that analysis:

1. A more accurate model for the chance for various PPG for the rest of the season

2. Accurate portrayal of variance of points for each position

These both are fine details overall, but that bit of extra accuracy is always helpful.

[TigerTedd]

Yeah but I think our chances of automatic is only about 10%.

I am not just concerned about Fester, Burnley, QPR, Reading and dare I say it Forest doing the double over us I worry about us losing to Blackburn, Leeds, Brighton etc, just as we did to Wigan . If we do , then with 22 games to go. we need 30 points say 10 wins for playoffs, or 9 wins and 3 draws- I think we will get that but only just and automatic is out of reach I think. Theres no wiggle room to allow for banana skins against the lesser teams.

To say you 'think it's more like a 10% chance' really does crap all over Albert's work. I don't know Albert, but clearly he's got some sort of background in statistical analysis. He's based his assertions on cold hard data going back for several seasons. What facts do you base your 10% on, can we see your working out?

Albert's already said that a 30% chance doesn't necessarily make it a likely result. In fact, it makes it an unlikely result. So what is the point if declaring it wrong and it should be 10% instead? It's still unlikely whichever way you look at it.

We all know that there is a very unpredictable nature to football, that's why we all love it. There's still a very small chance that we could get relegated. All you can do is come up with the odds of certain possibilities based on the data available, but bookies wouldn't exist if the odds weren't fallible.

[nfb]

Can someone summarise Albert's excellent post in two sentences please??

Erm, i'll try :- Maths is an exact science, football isnt !

Despite being battered by festa ,we still have a 95.7% of being in the play offs