Info and Instructions for the PlayoffComputer app

The basic purpose of the app is to take a league's info and remaining schedule to determine which teams have clinched a playoff spot, been eliminated, or are still in contention. Features include:

• "Paths-to-clinch" for eligible teams on what upcoming scenarios would result in clinching a playoff/division spot.
• Calculations for both division and wildcard spots.
• Odds for eligible teams on securing a playoff spot.
• Inclusion of many popular tiebreakers in the calculations.
• A quick "Monte Carlo" simulation for early season odds.
• New for 2020: Identification of teams that "control their own destiny"
• Data transfer from Yahoo, ESPN FF, My Fantasy League, and the companion Google Sheet.

Terms of Use

For the app to function properly, a league must:

• Have 4 to 16 teams in total, or in each conference.
• Play head-to-head games.
• Have 320 or less remaining games.
• For traditional leagues: Determine all, or all but one, playoff spots firstly by overall winning percentage (one spot can be reserved for team with highest points-for).
• For points-based leagues (e.g. hockey): Determine all playoff spots firstly by points-based with each win resulting in two or three points and each overtime loss resulting in one point.
• Have all pertinent data input into the app either thru data transfer or manual entry. Example: If the league uses head-to-head records as a tiebreaker, then the app requires past results.

• If transfering data soon after the end of the week, make sure that all data in the app is the most current, in particular that just-completed games do not show up in the "Remaining Games " section. ALL hosting sites have a lag between the conclusion of a week and all pertinent information being completely updated.
• The data transfer process can take several minutes.
• The name field made in calculations is limited to ten characters. It is strongly recommended for ease of reading the results that team names be manually adjusted to abbreviations or owner names.
• After data transfer, verify all data is complete and correct. The data transfer links can cut out on occasion.
• ESPN users: The app currently only accepts FF leagues for data transfer. The commissioner must set the "Make league viewable to public" setting to "yes".

• Your individual Google settings may dictate changes to these instructions.
• The generic PlayoffComputer sheet can be found here
• Begin by making your own copy of this spreadsheet (File - Make a Copy). This likely requires a Google account.
• See additional instructions on the Google Sheet. When ready to transfer data to the app, continue.
• Publish the spreadsheet copy (File - Publish to the web).
• Go to 'Sheets Home' (green box to the left of 'File'). Click or doubleclick to open the sheet.
• Using the URL link shown in the browser, cut and copy the spreadsheet KEY, which is the characters inside the full URL between the slash ( / ) after 'spreadsheets/d/' and before the following slash. Do not include the slashes.
• Paste the KEY into the web-app as shown within the "Google Sheets" data transfer option.
• Click "Import Sheet Data". "Authorize" if requested or if nothing happens after a minute.

• It is not absolutely necessary to enter W-L records, but if done for the sake of organization and having records shown in the results display, use the "T" column for overtime losses.
• Use the "PF" column for team points.
• The "PA" column can be used for goals or another stat if desired if needed for a further tiebreaker.
• See notes below for additional details.

• Verify data, especially the Remaining Games section, to make sure info is correct. This is the most common problem with data transfers from outside sources.
• Leagues with "bye-weeks" (rare) require one of the following if calculating "paths": 1) Ensure that the first listed games of any remaining week/round is not between teams that both had a bye the preceding week/round. manually adjust the listing of remaining games (e.g. switch placement of games within the grouping for the week/round); or 2) Use the "Non-standard schedule paths option" to select the number of games in the upcoming week/round to include in the "paths" analysis.
• "Points based league option": Select "yes" or "yes3" if the league is a hockey-type (wins = 2 or 3 points, overtime losse = 1 point). Also select "points" as the first tiebreaker.
• "Playoff berth information": If the league has divisions, select the number of division winner spots per division. This is almost always "1". For leagues without divisions select "0". Then select how many wildcard spots there are. For multi-division leagues, this is in addition to the total number of division spots (e.g. three-division league, six playoff spots, each division winner guaranteed a spot, resulting wildcard number would be "3"). For leagues without divisions, this is the total number of playoff spots. For leagues that award an additional playoff spot to the team leftover with the highest points total, only select the number of playoff spots in this section that are awarded by record.
• "Tiebreaker options": Select any tiebreakers for determining division winners (if applicable) and wildcard winners. See below "Tiebreaker info" section for important details and option descriptions. If head-to-head is a tiebreaker, the "Past Results" section must have the appropriate data. Hockey type points based leagues must have "points for" selected as the first tiebreaker.
• Except for points based leagues, a tiebreaker that cannot be firmly "projectable" such as points or all-play will be the last tiebreaker included in the calculations regardless if any are selected afterwards.
• "Points and All-Play tiebreaker options": The program will analyze the scoring of the league and, if set to "Automatic", calculate a reasonable amount of points that any one team can gain on another. "Most safe" increases the amount used in calculations that a team may gain, and should be used in leagues that tend to have a wide disrepancy in scoring possibilities. "Minimal changes" decreases the amount. "No changes" is unrealistic in most cases, but can be handy to run custom scenarios, particularly if the points tiebreaker is used for another stat. The same principle is used for the "All-Play" calculations.
• "Tied-game possiblity option": "Yes" will only be applied if there are 14 games or less remaining. "No" is recommended for most situations. For leagues in which a tie is a likely possibility it is recommended to select "No" until the final week/round, the results display can be almost unreadable with ties included in the calculations. Note if selected, this may cause the playoff odds probabilities displayed to be slightly incorrect
• "Doubleheader (two-round) paths option": For leagues that play such format each week, or as desired, but note that due to the increased number of games that have to be factored into the "paths" calculations, it is required that such "paths" analysis be limited.
• "Non-standard schedule paths option": For leagues that play a varying amount of games per each round, or if some results for the week/round are already entered, this can be used to search for "paths" for the specified number of remaining games.
• "Additional playoff-team by points option": For leagues that award a playoff spot to the highest scoring team leftover after the traditional w-l based calculations. Note the "paths" display may be confusing for some teams.

• "Monte Carlo": Thousands of random simulations of the remaining schedule are ran to provide a quick estimate of each team's playoff odds.
• "Basic": Calculations will be based on running every possible remaining scenario (if there are 22 or less remaining games and the "Tied-game possibility option" is set to "No", or 14 or less remaining games with it set to "yes") or a blend of possibilities with random simulations to determine which teams have clinched playoff spots, been eliminated, or their odds of making the playoffs. Additionally, the minimum number of wins necessary to gaurantee a playoff spot will be displayed, as well as whether those teams still in the running "control their own fate", i.e. are guaranteed a playoff spot by winning out.
• "Paths": Calculations will also include a listing of what scenarios for the upcoming week/round (or games, or two rounds, depending on options selected) would clinch a playoff spot for applicable teams. "Paths" will also be shown that would guarantee a tie for a playoff spot with others for leagues that use a points/all-play tiebreaker with a listing of which other teams each particular team might be fighting for as far as points are concerned.
• The "Basic" and "Paths" options can take many minutes, even hours, depending on many factors (league complexity, tiebreaker selection, number of teams still in contention, etc.).

• The tiebreaker calculation methods used may not exactly conform to the methods used by league hosting sites. Some known examples:
  • When a scenario is encountered in which three or more teams are tied and tiebreakers are calculated, the tiebreaker process starts over once any one team either secures or is eliminated from a spot in that scenario. It is believed, but not guaranteed, that this method matches with Yahoo, ESPN, and MFL.
  • When calculating tiebreakers for wildcard teams in multi-division leagues all tied teams are treated equally, a "best-in-division" rule is not employed. It is believed this method matches with Yahoo and ESPN but not with MFL.
  • Head-to-head methods do not match with the default setting in MFL, but do match with what many MFL commisioners set for their leagues.
• The following tiebreakers are available to be included in calculations:
  • "div pct": Division percentage, for multi-division leagues.
  • "conf pct": Conference percentage, for multi-conference leagues.
  • "points": Points-for. Higher trumps lower. As a tiebreaker that cannot be firmly projected, yet is one that will almost always resolve a tie at the end of the season, no further tiebreakers are accounted for in the calculations (with the exception of points based leagues) once this is employed.
  • "opp pts (hi)" or "lo": Points-against. "Hi" = higher trumps lower, "lo" = lower trumps higher. As with points, no further tiebreakers are accounted for in the calculations once this is employed.
  • "h2h #1": Head-to-head comparison that is only employed when two teams are tied. See below.
  • "h2h #2": Head-to-head comparison that is only applied when one team has swept the other team(s) that are tied.
  • "h2h #3": Head-to-head comparison that uses each team's winning percentage of games played against the other tied teams (believed to be default MFL setting).
  • "h2h #4": Head-to-head comparison that uses each team's number of wins against the other tied teams (believed to be default Fantrax setting).
  • "all play": All-play records. As with points, no further tiebreakers are accounted for in calculations once this is employed.
  • "non-div pct": Non-division percentage.
  • "random": This will randomly resolve ties. This is useful earlier in the season for leagues that use points or all-play as tiebreaker as it provides a simpler cleaner display result. To use, substitute 'random' in place of a points/all-play tiebreaker. NEW for 2020. Experimental but appears to work after brief testing. Of course, since this randomly chooses winners among tied teams for each scenario, it does not provide perfect results and should be avoided once there are fewer than about 25 games remaining.
• Based on best-known info, popular FF hosting sites and head-to-head tiebreaker recommendations for leagues which use head-to-head as a tiebreaker:
  • ESPN: Use "h2h #1" followed by "h2h #2" to get as close as possible to ESPN's methodology, which at last report is that if three or more teams are tied the h2h tiebreaker is only employed if they have played the same number of games against each other.
  • MFL: The default MFL H2H setting is flawed, if not manually adjusted use only "h2h #3". If the commissioner manually adjusts to comply with an NFL sweep-style rule, use "h2h #1" followed by "h2h #2".
  • Yahoo: Yahoo does not employ a H2H tiebreaker. Past results are not transferred automatically.
  • Fox: Use "h2h #1" followed by "h2h #2" to get as close as possible to Fox's methodology.
  • NFL Fantasy: Unknown.
  • CBS: Unknown.
• Points tiebreaker diatribe:
  • This discussion applies to all-play tiebreaker, if used, also.
  • It is almost universally agreed that points are a great way to break a tie at the end of the season. However, for the purpose of running scenarios to determine which teams have made the playoffs, which have been eliminated, and what needs to happen for teams to get in, the points tiebreaker presents an issue. The problem is points are impossible to predict and do not wholly relate to winning or losing. A fantasy team might win with 50 points while another might lose with 150 points at the same time. Thus, there is not a perfect way to project them in those calculations.
  • The program attempts to compensate by computing a realistic amount of points that one team could gain on another team over the remainder of the season. It will then only calculate whether a team has clinched or been eliminated in each scenario when a points advantage (or disadvantage) is too great to be overcome. It will then display the odds of each team finishing in a tie with others that will come down to the final points total at the end of the season, something that cannot be projected while the season is still ongoing.
  • There is one technical issue to this process when used on multi-division leagues that use points as a tiebreaker for both division winner and wildcard. If scenarios are encountered in which teams tie for a division with a worse record that the lowest wildcard team would have, requiring a points tiebreaker at the end of the season that cannot be ascertained currently, the wildcard odds displayed may be slightly off. The program does register when this possibility exists and will send a "weak division" message.

• The basic anlysis features the following:
  • At the top of the results message there will be wording as to whether every possible remaining scenario was computed or whether a probability based approach needed to be mixed in. There will also be a note on whether the possibility of tie-games was considered.
  • If there are any remaining scenarios in which a "weak" division winner cannot be projected due to unresolved tiebreakers (usually points) a message will appear stating that wildcard calculations may have been affected. This message will also appear in the extremely unlikely event that there are remaining scenarios in which absolutely no team could be determined at the present time to make the playoffs, but can be ignored in non-division leagues.
  • The number of wins the program determines will guarantee a wildcard spot will be shown as well as the minimum number of wins necessary to have any chance.
  • Each team will be listed with a description of their fate (e.g. clinched, eliminated, etc.). This message will vary depening on league structure (i.e. multi-division or not).
  • For teams that do not have a pre-determined fate, approximate odds will be listed for each team to claim a division (if applicable) and wildcard spot, or to be tied at the end of the season for such. The odds to "tie" represent the liklihood where a tiebreaker (usually points) whose resolution cannot yet be determined will have to break the tie at the end of the season.
  • The approximate odds for a particular team may add up to over 100% in multi-division calculations if there are scenarios in which a team may be in an unresolved tie for both the division and wildcard.
  • The approximate odds will be skewed if tie-games were factored in.
  • For the rare leagues that have an added "points-spot", the odds do not include calculations in which that spot cannot be determined yet.
  • Odds are rounded down (e.g. 98.85% = "98%") except for odds less than 1%, which will be shown with the appropriate number of decimal places.
  • If the points tiebreaker setting was changed to "Zero changes" (not recommended) the display will indicate that such advantages one team has over others may need to be maintained.
• The "paths-to-clinch" analysis features the following:
  • For multi-divisional leagues, the "paths" listings will be in two sections, division and wildcard.
  • For leagues with points tiebreakers, each section will then be split into two parts, one that shows ways a team can clinch a spot and one that shows ways a team can do no worse than end up in a tie for a spot. The other teams that this team may fight with will also be listed.
  • If the points tiebreaker setting was changed to "Zero changes" (not recommended), an asterisk following a "path" means that team must maintain a points advantage it holds over another.
  • If no "paths" are listed, no team can clinch the respective type of spot in the upcoming week/round (or two rounds or number of games as requested).
  • The "paths" feature has to be necessarily restricted in the interest of brevity and time in the following situations and may not be 100% complete:
    • If there are enough remaining games that a probability based approach had to be conducted.
    • The doubleheader option was selected.
    • If a custom number of games was selected that is more than one round's worth of games.

• The most common "issues" mentioned are:
  • ESPN users not having the league listed as "public" for the program to transfer data.
  • Users "requesting access" to the companion Google Sheet instead of making a copy as instructed.
  • Data be transferred immediately after the conclusion of MNF before the hosting sites have completely updated.
  • Discussions on the imperfections that points tiebreakers bring and the impossibility to firmly project a team's final points total in these calculations. The program's approach is the best possible method known to deal with this situation.
• Questions, suggestions, and compliments can be sent to "jomizec@gmail.com". Response time may be quite long.