代码改变世界

How to Construct the Input Bet String

2019-09-28 11:39  Dorisoy  阅读(310)  评论(0编辑  收藏  举报

The purpose of this section is to describe the format of the string which will submitted to the Python bet processor strats.py. This string is a series of lines that are Python variable declarative statements. There are a few things that readers not very with familiar with Python should know about:

  1. # starts a comment, so anything after and including # will be ignored, so lines starting with # can help make the intent of the declaration clear
  2. Blank lines are ignored. This may help the readability of the string when printed/debugged
  3. self. included in a variable declaration refers to the variable as a member of the class where it is declared. In our case it will be in a strats.Bet class object.

The rest of this manual will describe what variable to declare and what it means. These variables (attributes) will be introduced in the preferred order of declaration. Generally speaking, there is a default value for each variable, except for strats.Bet.bet_type and (currently) strats.Bet.nraces which must be specified.

[1] I would suggest that we let strats.py define strats.Bet.nraces from strats.Bet.bet_type.

Type of Bet Pool [Mandatory]

This is one of Quinella, Exacta, Trifecta, Pick 4, Pick 6, Quaddie, Superfecta. An example of this part of the string might be:

self.bet_type = "Quaddie"

Number of Races [Mandatory?]

This attribute is the number of races for which information is required. It is

  • 1 for Quinella, Exacta, Trifecta, Pick 4, and Pick 6
  • 4 for Quaddie
  • 6 for Superfecta

An example part of the string would be for the Quaddie:

self.nraces = 4
[2] strats.Bet.nraces should possibly removed from user input: but determined instead from strats.Bet.bet_type instead.

TAB Horse IDs [Mandatory]

Depending on the bet type, we will need either 1, 4, or 6 sets of horse IDs associated with the relevant race(s). To accomodate this the class variable strats.Bet.horse_ids is a list of lists. For a single race bet types this list only has one element: the list of TAB horse IDS for that race. As an example of the string for such a case is:

self.horse_ids.append([1,2,4,5,6,8,9])

Here is an example of the statements required for a Quaddie:

self.horse_ids.append([1,2,3,4,5,6,7,8,9,10,11])
self.horse_ids.append([1,2,3,4,5,6,7])
self.horse_ids.append([1,2,3,4,5,6,7,8,9])
self.horse_ids.append([1,2,3,4,5,6,7])

TAB Odds [Mandatory?]

Either the TAB Odds, User Odds, or both must be specified. Depending on the bet type, we will need either 1, 4, or 6 sets of odds associated with the relevant race(s). To accomodate this the class variable strats.Bet.tab_odds is a list of lists. For a single race bet types this list only has one element: the list of TAB horse IDS for that race. As an example of the string for the single race case is:

self.tab_odds.append([2.0,4.2,4.2,5.5,5.9,"-",11.0])

Any entry of a non-numeric character indicates a scratching for that horse.

Here is an example for a Quaddie:

self.tab_odds.append([5.2,3.2,16.2,5.8,9.3,34.8,6.7,9.7,14.3,10.5,60.])
self.tab_odds.append([1.3,10.0,12.0,3.1,15.1,1.6,2.3])
self.tab_odds.append([8.3,31.0,4.2,10.0,6.0,7.2,3.2,20,42.6])
self.tab_odds.append([31.7,13.7,4.7,22.6,11.7,2.4,11.7])

User Odds [Mandatory?]

Either the TAB Odds, User Odds, or both must be specified. Depending on the bet type, we will need either 1, 4, or 6 sets of odds associated with the relevant race(s). To accomodate this the class variable strats.Bet.user_odds is a list of lists. For a single race bet types this list only has one element: the list of TAB horse IDS for that race. As an example of the string for the single race case is:

self.user_odds.append([2.0,4.2,4.2,5.5,5.9,"-",11.0])

Any entry of a non-numeric character indicates a scratching for that horse.

Here is an example for a Quaddie:

self.user_odds.append([5.2,3.2,16.2,5.8,9.3,34.8,6.7,9.7,14.3,10.5,60.])
self.user_odds.append([1.3,10.0,12.0,3.1,15.1,1.6,2.3])
self.user_odds.append([8.3,31.0,4.2,10.0,6.0,7.2,3.2,20,42.6])
self.user_odds.append([31.7,13.7,4.7,22.6,11.7,2.4,11.7])

Aggregation of IDs and Odds

It may be more convenient (and doesn’t affect the bet logic) if the TAB IDs and odds (user and/or TAB) are grouped together:

self.horse_ids.append([1,2,3,4,5,6,7,8,9,10,11])
self.tab_odds.append([5.2,3.2,16.2,5.8,9.3,34.8,6.7,9.7,14.3,10.5,60.])
self.user_odds.append([4.1, 3.8, 19.1, 8.0, 7.6, 35.8, 7.2, 11.1, 11.6, 10.9, 56.8])

self.horse_ids.append([1,2,3,4,5,6,7])
self.tab_odds.append([1.3,10.0,12.0,3.1,15.1,1.6,2.3])
self.user_odds.append([1.6, 13.0, 7.2, 2.8, 20.8, 2.3, 1.9])

self.horse_ids.append([1,2,3,4,5,6,7,8,9])
self.tab_odds.append([8.3,31.0,4.2,10.0,6.0,7.2,3.2,20,42.6])
self.user_odds.append([5.0, 34.8, 5.3, 14.3, 4.3, 8.5, 3.2, 28.9, 61.3])

self.horse_ids.append([1,2,3,4,5,6,7])
self.tab_odds.append([31.7,13.7,4.7,22.6,11.7,2.4,11.7])
self.user_odds.append([29.3, 13.9, 4.7, 26.1, 15.2, 1.5, 8.9])

Jockeys [Option]

For multi-race betting (Quaddie and Superfecta) jockey information can be input. (Jockey information for single race betting pools is forbidden.) An example input for a Quaddie is kept in strats.Bet.jockey_list which is class attribute of a list of lists:

self.jockey_list.append([1,2,3,4,5,6,7,8,9,10])
self.jockey_list.append([11,21,13,41,51,61,17])
self.jockey_list.append([11,21,13,41,51,61,81])
self.jockey_list.append([431,351,613,137,831])

If jockey info is provided for one race, it needs to be provided for all. If provided, the jockey information needs to be provided for each runner in each race. If jockey information is not provided, then strats.Bet.jockey_list defaults to the empty list, and jockey information plays no part in bet discrimination.

Trainers [Option]

For multi-race betting (Quaddie and Superfecta) trainer information can be input. (trainer information for single race betting pools is forbidden.) An example input for a Quaddie is kept in strats.Bet.trainer_list which is class attribute of a list of lists:

self.trainer_list.append([6,2,3,4,5,6,7,8,7,60])
self.trainer_list.append([66,26,63,46,56,66,67])
self.trainer_list.append([66,26,63,46,56,66,86])
self.trainer_list.append([436,356,663,637,836])

If trainer info is provided for one race, it needs to be provided for all. If provided, the trainer information needs to be provided for each runner in each race. If trainer information is not provided, then strats.Bet.trainer_list defaults to the empty list, and trainer information plays no part in bet discrimination.

Scaling Odds [Option]

The attributes, strats.Bet.scale_tab and strats.Bet.scale_tab, are parameters that are used to scale the odds data. They affect scaling according to:

  • if = -1, the odds are unscaled: the probabilites are the inverse of the win price line
  • if = 0, the odds are normlized: the probabilities of the price line add up to 1
  • if between 0 and 100 (% of normal), the odds are scaled by odds*scale_factor/100

Here is an example:

self.scale_tab = -1 # unscaled
self.scale_user = 85 # scaled by 85% to try to approximate TAB takeout

By default the scaling of the odds is normalized.

Standouts [Option]

For single race pools, the user can stand out any place in the combination The relevant attribute associated with each place is given by standout_x, where x is the place. For example if we wished to stand out horse 1 and 2 for first, 1,2, and 7 for second, and the field for third in a trifecta bet, the string involved would be:

self.standout_1 = [1,2]
self.standout_2 = [1,2,7]
self.standout_3 = []

The third line would not be necessary as the field is the default for the standouts.

For the multi-race pools, the standouts are for the winners of each race, so standout_x is the stadout for the xth race in the Quaddie or Superfecta. An example string standout element for a Quaddie bet might be:

self.standout_1 = [1,4]
self.standout_2 = [2,3,7]
self.standout_3 = [1,5,9,11]
self.standout_4 = [2,8]

Filter Price Line by Interval [Option]

We can filter the win price line by specifying the attributes strats.Bet.taboddsx and/or strats.Bet.useroddsx where x refers to an odds place restriction for a single race bet, or to a odds win odds restriction for a multi race bet. These restriction are expressed as an odds minimum, odds maximum tuple. For example, if one was very confident of the user win price line for a race, one could express the restriction for a trifecta bet as:

self.userodds1 = (1,4)
self.userodds2 = (1,10)
self.userodds3 = (8,20)

As another example, if one had a strong belief that runners were only in the market if the TAB odds wre less than 10, we could put this restriction on a Superfecta bet:

self.tabodds1 = (1,10)
self.tabodds2 = (1,10)
self.tabodds3 = (1,10)
self.tabodds4 = (1,10)
self.tabodds5 = (1,10)
self.tabodds6 = (1,10)

Restriction on Exotic Odds Range [Option]

After the TAB and/or user exotic odds have been calculated from their respective win price lines using the rules above, one may decide to restrict the bets to a particular exotic odds range (min,max). This is done by specifying the attributes strats.Bet.tab_div_range and/or strats.Bet.user_div_range. For example, suppose in trifecta betting we want to restrict the calculated user trifecta odds to less than 50 and the calculated TAB odds to great than 50. We might specify this condition as:

self.user_div_range = (1,50)
self.tab_div_range = (50,1000000)

If there is to be on restriction on the exotic odds range, then these attributes are set to False, which is the default.

Overlay (Option)

If set to True, then the overlay attribute, strats.Bet.overlay, then combinations will be accept only if the TAB exotic odds are at least strats.Bet.overlay_var per cent greater than the user exotic odds. Here is an example where the user odds are required to be 150% better than the TAB odds:

self.overlay = True
self.overlay_var = 150

These attributes are only relevant if both win price lines are specified.

Top Combinations by Probability (Option)

If the attribute strats.Bet.percent_by_number is given, then the top % combinations specified are accepted. If there is only one win price line specified, then the combinations are sorted by that price line. If both price lines are specified, then they are rated by the user price line. For example, suppose we only want the top 25 per cent of combinations for a given set of exotic combination as calculated from the rules above. The relevant string would be:

self.percent_by_number = 25

Top Combinations by Value (Option)

If the attribute strats.Bet.percent_by_value is given, then the top % combinations rated by value are accepted. Value is defined as the ratio of the TAB odds to the user odds. Clearly, both price lines need to be defined for this option to be invoked. For example, suppose we only want the 10 per cent of combinations for a given set of exotic combination as calculated from the rules above. The relevant string would be:

self.percent_by_value = 10

Staking (Option)

One can specify either flat staking or staking based on taking out a specified amount of money. Which scheme is used depends on the Boolean attribute strats.Bet.flat_stakes. If True (the default with stakes/combination of 0.50), then one could specify the flat staking at $1 per combination as:

self.flat_stakes = True
self.stakes = 1.0

Alternative, one could weight the stakes so as to take out a fixed amount of winnings depending upon the odds. If both price lines have been input, it will be necessary to specify which price line will be used for the odds. This is in the attribute strats.Bet.takeout_priceline which can be either “tab” or “user”. Here is an example takeout scheme using the TAB odds to take out $500 per combination:

self.flat_stakes = False
self.takeout_priceline = "tab"
self.takeout = 500

Rounding (Option)

If not flat staking, the individual bet/combination outlay can be rounded in various ways depending upon the strats.Bet.rounding attribute:

  • “Exact” - as calculated
  • “Round Up” - round up to nearest .50 increment (default)
  • “Round Down” - round down to the nearest .50 increment
  • “Nearest” - round to the nearest .50 increment

An example of rounding to the nearest $.50 increment is:

self.rounding = "Nearest"

Total Outlay (Option)

After all of the other rules are in place, the total outlay for the bet can be capped with the attribute strats.Bet.outlay_amount. This is currently set by default to a million dollars. Here is a specification that puts on a more reasonable cap:

self.outlay_amount = 100