Tiers?

[borodin]

I'm reading the issues of the Follow Up. In issue 75, it says that tiers are not the same as ranks and that they only deal with odds. It also talks about tiers being horizontal rankings.

I couldn't find an answer in a forum search.

Could somebody explain exactly what tiers are?

[Bill V.]

Hi

Tiers are different than ranks.

When we rank something 1 is always better than 2,
2 is always better than 3, 3 is better than 4 and so on

When Doc spoke about tiers being horizontal he meant that
based on his readouts1 is just as good a 2 or 3
and on occasion 4. The difference being the final odds

example
Horse 1 is 5/1
Horse 2 is 5/2
Horse 3 is 10/1

At one time when the Methodology ranked horses the bets would be
Horse 1 and 2 full win bets and then bet horse 3 with a separate bank.
(from the yellow manual)

Later Doc changed and said to think in terms of tiers.
The tiers of 1 2 and 3 meant the horses were all equal in ability.
They were all in the same group but that since horse 2 met his minimum odds
requirement, and horse 3 was the best value those were to be his suggested bets.

Rankings means right or wrong - better or worse

When I think of tiers I think of going to Lincoln Financial field in Philly to watch the Might Eagles play
Even though I'm sitting in tier 3. I'm in the same wonderful place as those people in tier 1 and 2

Im still in the same place as # 1 and #2

Thanks
Bill

[borodin]

Thanks for the reply.

I'm still missing it. Does this mean that we have to decide which horses are roughly equal in ability and separate them by tier, or do the top 3 always count as being equal in ability and in the same tier?

[Because I Can Jim]

Hey borodin,

First off, this is only an example. It is not meant to imply paceline selection, contender selection, etc... It is not about handicapping. It is only meant to discuss ranks vs tiers. I did not include the track, date, etc... because I don't really care who won the race or at what odds.

So with that said,....

Think of ranks as first best, second best, third best, etc....

Based upon this eample, H4 would be ranked 1st tied with H7.
H6 would be ranked 2nd.
H5 would be ranked 3rd.
H1 would be ranked 4th.

Simple enough. You already know that.

Looking at the sample image, let's assume that the M/L odds are the odds you would be getting on the tote board with 30 seconds to post and you want to make a wager.

As Bill said "At one time when the Methodology ranked horses the bets would be..." on the H4 and H7 depending on your minimum acceptable odds wagering strategy. Assuming you accepted 2-1 as your low odds horse, you would wager 60% of your wager on the H7 because it is the low odds horse and 40% of your wager on the H4 horse. This is because they are ranked tied for 1st. If there wasn't a tie, you would wager on the horses ranked 1st and 2nd, again, using whatever you accepted as your minimum odds.

Now on to tiers.

When ranking horses, you are trying to say "I think H4 and H7 are going to win and therefore, I rank them first. I can't distinguish between the two so I rank them both 1st. I also think H6 has the next best chance of winning "if" H4 and H7 do not win; therefore, I am ranking it 2nd."

As Bill also pointed out "Later Doc changed and said to think in terms of tiers."

Think of it this way....

There is a photo finish between three horses. Horse A beats Horse B by a nose who finishes a head in front of Horse C. At the finish, Horse A is ranked 1st, Horse B is ranked 2nd and Horse C is ranked third.

However....

As they came down the stretch and just at the finish (as Howard put it) you blinked your eyes, you would not know who finished 1st, 2nd, or 3rd. It was that close. Can you really say that Horse A is better than B or C, or that Horse B is better than A or C and lastly Horse C is better than A or B? If they ran the race again and again and again, would the order of finish remain the same beaten exactly the same way?

This is where thinking in terms of tiers comes in.

Pre-race, you could have used whatever handicapping method you chose to select contenders and pacelines and rank the horses in the order that you think has the best chance to win.

Or...

You can think in terms of tiers....

You can select pacelines and contenders and look at the BLBL tab (letting the computer algorithms work their magic), but, instead of thinking in terms of rank, you can think - here are my top 5 contenders. I know that some of the horses are better than others, but, I am more concerned with the "price" I am getting on the horses (overlays) than I am concerned with how I think they are ranked.

For example, Horse 4 is going off at odds of 8-1 and H7 is going off at odds of 2-1. They both are tiered the same for all intents and purposes. Which would I wager on? H4! I'm going to get a much better price if it wins versus the H7. Remember, when I blink my eyes...who knows which would finish first at the wire?

So far so good, yes?

Ok. This is one isolated race. If you model your wagering decisions on a Wagering Decision Form, you will find out based on "your" handicapping (i.e. selecting pacelines, contenders, etc...) what your percentages of winners are in the different tiers.

For example, lets say your percentage of winners by tiers looks like this:
Tier 1 = 40%
Tier 2 = 19%
Tier 3 = 6 %
Tier 4 = 14%

Now when you are looking at the horses, you would have a tendency to lean more towards the Tier 1,2 and 4 or in this case H4, H7, H6 and H1.
H5 would not be considered even though it is tiered 3 for win unless the odds were so high that it would warrant a small wager, but in this case it is going off at 3-1. Based upon your wagering decisions, H5 (a Tier 3 horse) only has a 6% chance of winning. Would you wager on it? Probably not, not at 3-1, but, at 30-1, maybe...

So, let's say your minimum acceptable odds are 5/2-1...
That leave horses 4, 6, and 1 to be considered for wagering as a function of your wagering strategy for win.

Notice H7 is not a bet. Yes, it is ranked tied for 1st, but, it is not "a blind bet" just because it looks like it is "ranked" 1st.

To give a starker contrast and to drive home the point further.... let's flip the percentages around...

Based upon your wagering decision form:
Tier 1 = 14%
Tier 2 = 6%
Tier 3 = 19%
Tier 4 = 40%

Since we are looking at the information on the BLBL tab by tier and not rank, I would have more of a tendency to wager on the H1 and H5 if they met my minimum acceptable odds criteria which they do at 5/2-1. The wager would be 60% on H5 and 40% on H1 (remaining consistant with our example, but, if you use 50/50 or whatever, so be it).

The thing here is that we are not concerned with which horse is best. We are concerned with the price we get for winners based upon which tiers we find our winning horses are coming from which is based upon our handicapping. Using these percentages, I would have a 59% probability of having the winner (40% + 19% = 59%). If I am getting the right price, I cannot help but make money in the long run. This is because I know my percentage of winners by tier that I am getting and at what price.

Lastly, let me change this up just a bit. Let's say that my minimum acceptable odds are 4-1....

In the first example set of Tier percentages, I would have a wager on H4 with my 40% win percentage from Tier 1. I would not wager on H7, H6, H5 because they do not meet my minimum 4-1 odds criteria. I would need a lot more than 6-1 on H1 because my percentage is only 14% in Tier 4. OR I would PASS the race if I was not comfortable wagering on one horse.

NOTE: The Tier 1 is 40%. However, that does not mean that H4 as a single horse wager is 40%. It would be less than 40% due to ties being considered as "one" horse when either of the horses win when entering on the wagering decision form. However, in this example, at 8-1, it would be worth a wager.

With the second example set of Tier percentages, the situation would be the same. The wager would be on H1 (Tier 4) or a pass for the race unless the odds for H4 are large enough. The difference here would be that I would be much more confident that as a single horse wager, H1 would win much closer to the 40% of the time.

I hope that helps. A long explanation, I know...., but, I don't seem to quite know how to keep things short!

Best of luck.
Jim

[Ted Craven]

Originally Posted by Bill V.

When I think of tiers I think of going to Lincoln Financial field in Philly to watch the Might Eagles play
Even though I'm sitting in tier 3. I'm in the same wonderful place as those people in tier 1 and 2

You know Bill, that's a GREAT visual metaphor for re-thinking how to view the upper ranks of the Bottom Line/Betting Line (BL/BL) screen - one which I never thought of before, but it makes perfect sense for Sartin's concept of 'tiers' versus 'ranks'. He thought that the 'western' mind was very much (and overly) rank oriented (or perhaps 'absolute-oriented') - who's absolutely number #1 - perhaps from our orientation to sporting event winners (#1), or politics (who out-ranked/out-polled whom), etc. The difference between something like tiers is much less absolute than the difference between the concept of ranks – at least in some people’s thinking. Sartin was always challenging people’s default thinking patterns.

What kind of a baseball league would it be if the top 3 ranked teams were all considered of equal ability before every game? But in handicapping horse races, we're not interested in who wins, per se - we're really interested in how much we get paid over a series of races!

Sartin mostly used the term 'tier' in the context of the Odds Line - the BL/BL. There are absolute 'ranks' 1-2-3-4-5 (that's how the screen is sorted) - but if you consider that, say, the top 3 of those ranks stand an equal chance to win over a series of races, then in a given race the horses we're most interested in are those who will pay us the best.

Another visual metaphor (for me) comes from a Riverdance concert me and my wife got last minute rush seats for once in Toronto. One seat was in the front row (i.e. 'Tier #1') and the other was 10 rows back (say, Tier #3). My wise wife gave me the Tier #1 and she took Tier #3. At the end of the (remarkable) concert, I had a stiff neck from staring up and was covered in sweat and spittle from the dancers 3 feet away. She, on the other hand, was rapturously tapping a tune and in Celtic heaven. Who got the better Tier? We both had a good time, a great view - but perhaps different value ...

The low odds favourite is often in rank (Tier) #1 - loses 67% of the time and pays insufficiently. The rank(Tier) 2 horse wins less and pays more, and the #3 horse wins even less and pays even more. It behooves us to think of these 3 ranks as seating tiers to an event (game/arts event) where the top 3 tiers are all just as good as the other (as Bill said), except that there's some usually publicly visible attributes which result in lower pricing of the higher ranked horses.

By letting your mind relax in terms of 'who's number #1' and accepting #2 and #3 (and #4 if some factors look right - let's discuss 'Binder improvers' later) - you can put yourself on 2 horses out of your Top 3 which pay out often enough, at sufficient net odds (i.e. for the 2 bets) to make a profit over a series of races. That's the 'tiers versus ranks' thesis.

Borodin - The BL/Bl ranks arise on that screen based on the paceline(s) you choose for the real contenders. You can either choose each line yourself, or let the software do it for you using the Paceline Selection Tool (e.g. Best of last 3), or a combination. This process results in a ranking of horses on the BL screen in the Analysis panel.

You don't have to make judgements about which horse are equal to each other - other than by looking at that BL screen (after you have qualified the horses as contenders to perform their best again today). Then you pick horses according to their offered odds at bet-time.

An example always adds clarity to a theoretical discussion! I was responding to another race query by Bill last evening, so I'll just use that one:





Let's use final odds as stand-ins for 'bet-time' odds. Let's consider the top 3 ranks/tiers: 8-4-2.

#8 = 1.7
#4 = 2.6
#2 = 4.6

Imagine each horse is sitting on a broad tier (imagine a stadium, with a game down on the field), and beside those numbers sit 'friends' to the left and the right: other ranks from the Primary Factor group, the Supplementary factor group, the Total Energy, the VDC, the CSR, etc). Each of the 3 tiers have an excellent view of the field, though 'technically' the lower one (Tier 1) is absolutely closer - but who cares.

Imagine each has an equal chance to win. Betting on the Top tier/rank will win you ~ 33 - 35% of the time, and possibly pay you too little to profit over a series of races (depending on how strong that top status is - there are other ways to qualify that). The 2nd tier/rank will win less than the 1st, and the 3rd less than the 2nd, with each paying (usually) increasingly higher average (Win) mutuels.

In the above images, note that the #2 and #4 are effectively on the same 'tier'. The gap between the BL #s (20.8 vs 20.3) is effectively nil - as suggested by their identical Odds Line value (1.8 each)

The strategy Sartin is proposing is to consider if there is a sound wager from those top 3, using 2 horses if possible. Here, you could skip the top ranked #8 and seek a bet from the #4 and #2, getting a net odds range here of 1.3 ... 2.3 and having 2 chances to win (divide each odds in half for 2 bets). Many would not accept this low of net odds, but then you have 2 chances to win.

Sartin labelled this concept as 'wagercapping'. Search this site (the Google Site Search tool) or review numerous Follow Up Issues for more discussion and illustration in his words.

So - 'tiers' encourage you to loosen hold on the notion of absolute ranks, and seek value from among the upper echelons of how the software has comparatively rated the horses.

We have not touched in this discussion on the crucial topics of Contender identification and Paceline selection - only wager assessment from a summary handicapping screen including 'vetted' contenders.

Hope this helps!

Ted

[Ted Craven]

Jim,

Thanks for your explanation too! (I started mine last night and just finished it now ...). We covered many of the same points!

cheers,

Ted

[Because I Can Jim]

Ted,

You are welcome. I am glad to see I am not the only one who writes a lot! LOL

Best,

Jim

[borodin]

I think I have it now. Thanks to all of you for clearing it up.