Are There Better Ways To Limit Team Strength?

[Grachan 7,362]

I'm wondering about this. Remember the old days of rider control? Averages weer really a good old nerdy guide to form in those days.

 

I'm not sure that rider control would work any more, but if averages were taken out of the equation, Swindon would probably have been able to sign Darcy Ward without relleasing Kyle Howarth.

 

Also, even if they did, coventry probably wouldn't be able to sign him because they are top of the league.

 

Yet, because team strengths are governed by averages, both could happen.

 

Certainly, if team strengths were governed by a committee, it would have to be a neutral one, not the BSPA.

 

I guess averages are so ingrained in the psyche of team building nowadays that people would never accept uneven averaged sides.

 

I have a programme somewhere from 1976 that showed the averages of the sides at the start of the season. The highest was around 55 and the lowest about 39. Nobody gave it much thougt then.

 

What other methods could there be to ensure even balanced teams?

[SCB 0]

I have a programme somewhere from 1976 that showed the averages of the sides at the start of the season. The highest was around 55 and the lowest about 39. Nobody gave it much thougt then.

Technically, you could get very close to that 39 point team being the better team of the two in theory! But only with bonus points and a heat 15. 3-3 in every race where one team is always first and last, there top two go into heat 15 on maximums but one of them drops a points, they score 46 but have a combined team average of 40.2. But the other team scores 44 +14 bonus points and average 53.6

 

Don't you just love averages!

[ouch 1,191]

Rider control, if it was run like it was in the 70's we'd have to lose Zagar and Coventry would lose Robson.

[SpecialOne 68]

I don't think much can be done, manipulation of averages will always go on in the current system as it has been watered down so much and the team limit average is so low. It wasn't that long ago that teams would try and a good rider on a low average to be reserve and take 7 rides, think likes of Chrzanowski at Swindon a few years back. Now this is limited by the draft as most riders in it are not going to be chosen to replace a rider in the main body of the team (unless that rider is Miedzinski). So the manipulation of averages will only go on more to fit these vast improvers in the main body of the team for the following year or to create space for a new rider this season.

 

All it does it further water down the product we get as fans.

[oldace 1,678]

[quote name="grachan" post="2642971" timestamp="1436281411

 

I have a programme somewhere from 1976 that showed the averages of the sides at the start of the season. The highest was around 55 and the lowest about 39. Nobody gave it much thougt then.

 

What other methods could there be to ensure even balanced teams?

 

And with an average of just under 10 bonus points a match, just under 5 per team, your 55 point team is 50 and your 39 point team is 34 using the modern way of not including bonus points leaving a mean team average around 42, all exactly as it would work today.

[Grachan 7,362]

And with an average of just under 10 bonus points a match, just under 5 per team, your 55 point team is 50 and your 39 point team is 34 using the modern way of not including bonus points leaving a mean team average around 42, all exactly as it would work today.

Hate to be a smart arse but that's actually the median, not the mean. :-)

 

But yes, you're right. It will always, inevitably, work out at around the 42 mark when all 7 get an average.

[waiheke1 4,295]

No that's the mean.

Though with a population of two there is an argument it is also the median, being the midpoint between the two middle numbers.

[Grachan 7,362]

No that's the mean.

Though with a population of two there is an argument it is also the median, being the midpoint between the two middle numbers.

He quoted the highest and the lowest numbers and took the middle. That's the median. There were about 18 teams in the league. Had there been 2 teams it would also have been the mean.

[SCB 0]

He quoted the highest and the lowest numbers and took the middle. That's the median. There were about 18 teams in the league. Had there been 2 teams it would also have been the mean.

The median is where you line up all numbers in order and take the middle value. You need an odd number of values to get a real median. If you had numbers 7, 1 and 19, the median is 7 as it's the middle of the 3 numbers. But the mean would be the sum of all 3 divided by 3 - 9 in my example. Edited July 7, 2015 by SCB

[Grachan 7,362]

So what is it if you take the highest number and the lowest number and take a midpoint? Is there a name for that? We need a name for that.

 

Edit. Just googled it. It's the range. Pretty obvious when you think about it.

Edited July 7, 2015 by grachan

[waiheke1 4,295]

He quoted the highest and the lowest numbers and took the middle. That's the median. There were about 18 teams in the league. Had there been 2 teams it would also have been the mean.

surely he took the highest and lowest and divided by two? To know the median he would have needed the numbers in between

So what is it if you take the highest number and the lowest number and take a midpoint? Is there a name for that? We need a name for that.

 

Edit. Just googled it. It's the range. Pretty obvious when you think about it.

The range is the highest number less the lowest number, nothing to do with a midpoint?

[Gemini 4,894]

Blimey! I'm glad I'm not interested in averages and statistics as I'm afraid you've all lost me now. Median, mean, range?

[Grachan 7,362]

Just to confuse Gemini even further, I can now confirm that the mean of the highest and lowest numbers is called ...

 

Da da daaaaaaaaaaaaaaa..........

 

The mid-range.

 

No I'm right this time. I really am. I double checked.

[Humphrey Appleby 13,960]

What other methods could there be to ensure even balanced teams?

You don't want to have balanced teams, and it's not achievable as demonstrated by the fact that you still get teams at the top and bottom of the leagues regardless of how punitive the points limit is. What you want to avoid is complete mismatches and years of dominance by a single team, whilst providing some encouragement for teams to develop their sides.

 

Any system must allow teams to develop over 2 or 3 seasons without fear of being broken-up if they merely finish above mid-table, allow them to genuinely develop without needing to resort to average fiddling (e.g. through manipulated assessed averages), and protect young riders from the vagaries of team equalisation. Equally though, teams shouldn't be allowed to strengthen indefinitely, and there should be an upper limit whereby they have to relinquish a rider or riders if they become too dominant.

 

I think it's difficult to get away from a system based on averages, although some sort of draft system might be considered whereby teams are allowed to retain a certain number of riders at the end of each season, with the rest released into a pool along with any new riders, whereby they're picked by teams in reverse finishing order. However, a draft system only really works well where there's a single dominant league that everyone aspires to ride in, and that doesn't really happen in speedway.

 

The other oft-suggested alternative is grading, but that's just crude averages and makes riders near the top of their grades more in demand than ones at the bottom. For me it's no improvement on the points limit.

 

So I'd introduce an a variable points limit that allows teams to build to different limits depending on how many riders they retain from the previous season. The basic points limit would be set around the break even point (i.e. 42 points) to which every team would be allowed to build, but teams would be allowed to build to a higher level depending on how many riders they retained from the previous season. A suitable tariff would need to be worked-out, but let's say 0.5 points per rider, which would in principle allow a team to build up to a maximum of 45.5 points.

 

The idea behind this is that it would prevent teams from simply going-out and signing a strong side, but it would allow up-and-coming teams to develop over two or three years without being penalised. Teams only finishing mid-table would be able to keep their sides together as it would be unlikely they'd improve by more than a couple of points in a season, whilst the upper limit (45.5) would still prevent successful teams from becoming overly dominant. It should also prevent middle- and lower-order riders who improve their averages, from being squeezed out of teams as their continued presence would allow a higher points total. Equally, this might encourage more rider loyalty as it would be advantageous average-wise for riders to stay with the same team.

 

Combined with this should be an incentive for teams to develop new riders. So, I'd allow U21 riders or those in the first three years of their careers to ride on a low fixed average (e.g. 3.00), but only if they stay with the same team. Having said this, they'd be allowed to be loaned to one team in another league (either upwards or downwards) for a maximum of (say) one or two seasons to gain experience without losing their 'junior' status.

 

Average manipulation is almost impossible to entirely eradicate, but could be mitigated by using factored averages. In other words, a rider's team building average would be comprised of two components - their starting average and their actual average for the season. At the beginning of the season, the starting average would be weighted more heavily, but towards the end of the season their actual average would be more significant. The weighting would depend on how many matches had been ridden, so if a rider managed to ride in all league matches, their team building average would be their actual average. The factored average at the end of the season would then constitute the starting average for the following season.

Technically, you could get very close to that 39 point team being the better team of the two in theory! But only with bonus points and a heat 15. 3-3 in every race where one team is always first and last, there top two go into heat 15 on maximums but one of them drops a points, they score 46 but have a combined team average of 40.2. But the other team scores 44 +14 bonus points and average 53.6

Don't you just love averages!

The lower averaged team could be the better team, but I don't think you can get close to the 39 points using the 13 heat formula used in 1976. So 12 x 3 (no bonus points scored) + 4 = 40 which is a combined CMA of 43.08, and 12 x 3 (with 12 bonus points) + 2 = 50 which is a combined CMA of 53.85.

Edited July 8, 2015 by Humphrey Appleby

[Truro Robin 39]

I think that in an ideal world, whichever team finishes with the highest combined average, that figure should be the points limit for the following season.

 

If you look at he latest Green Sheet figures, there is only 3.18 points between the highest ( Swindon 34.71 ) and the lowest ( Leicester 31.53 ) so you don't have a huge disparity to deal with in the first place.

 

Obviously, if you wanted to do away with doubling up and the Fast Track reserves, then both the Elite League and Premier League would need to lower their team building figure for a season to allow for this to happen, but that's another story!