For games in different competitions, a 1 means the first half and the final score had the same result (win, draw, or loss for the home club ignoring the margin of victory or defeat) and a 2 means the second half and the final score had the same result (both could happen and N means neither happened): UEFA Champions League 2010-2011 Group Stage Matchday 5 today: Spartak Moscow-Marseille: 1 and 2 Sporting Braga-Arsenal: 2 Chelsea-MSK Zilina: 2 AS Roma-Bayern Munich: 2 Ajax-Real Madrid: 1 and 2 Auxerre-AC Milan: 2 Basel-CFR Cluj: 1 Partizan-Shakhtar Donetsk: 2 Total: Neither happened 0 times 1 only happened 1 time 2 only happened 5 times 1 and 2 happened 2 times 1 with or without 2 happened 3 times 2 with or without 1 happened 7 times English Premier League November 20-22, 2010 (ten games, each club played exactly once): Arsenal-Tottenham: 2 Birmingham City-Chelsea: 1 Blackpool-Wolverhampton: 1 Bolton-Newcastle United: 1 and 2 Manchester United-Wigan Athletic: 1 and 2 West Bromwich Albion-Stoke City: 2 Liverpool-West Ham United: 1 Blackburn-Aston Villa: 1 and 2 Fulham-Manchester City: 1 Sunderland-Everton: 1 and 2 Total: Neither happened 0 times 1 only happened 4 times 2 only happened 2 times 1 and 2 happened 4 times 1 with or without 2 happened 8 times 2 with or without 1 happened 6 times 2 happened more frequently than 1 in the Champions League, while 1 happened more frequently than 2 in the Premier League. Because the second half has more stoppage time, with a large sample I would expect 2 to happen a little more often than 1. I expect to post more in this topic later.
Major League Soccer 2010: First game for each of the 16 clubs on March 25-27 Seattle-Philadelphia: 1 Chivas USA-Colorado: 2 FC Dallas-Houston: 1 and 2 Columbus-Toronto FC: 1 and 2 New York-Chicago: 1 Kansas City-D.C.: 1 and 2 San Jose-Real Salt Lake: 1 and 2 Los Angeles-New England: 1 Total: Neither happened 0 times 1 only happened 3 times 2 only happened 1 time 1 and 2 happened 4 times 1 with or without 2 happened 7 times 2 with or without 1 happened 5 times World Cup 2010: Round of 16 (overtime and penalty kicks ignored, First Half result and Second Half result are each compared to the result after regulation): Uruguay-South Korea: 1 United States-Ghana: N Germany-England: 1 and 2 Argentina-Mexico: 1 Netherlands-Slovakia: 1 Brazil-Chile: 1 and 2 Paraguay-Japan: 1 and 2 (scoreless after regulation) Spain-Portugal: 2 Total: Neither happened 1 time 1 only happened 3 times 2 only happened 1 time 1 and 2 happened 3 times 1 with or without 2 happened 6 times 2 with or without 1 happened 4 times
World Cup 2010 Qualifying: CONCACAF Semifinal Group 1 (August 20, 2008-November 19, 2008) Cuba-Trinidad and Tobago: 1 and 2 Guatemala-United States: 2 Trinidad and Tobago-Guatemala: 1 and 2 Cuba-United States: 1 United States-Trinidad and Tobago: 1 and 2 Guatemala-Cuba: 2 United States-Cuba: 1 and 2 Guatemala-Trinidad and Tobago: 1 and 2 (scoreless draw) Cuba-Guatemala: 1 Trinidad and Tobago-United States: 2 Trinidad and Tobago-Cuba: 2 United States-Guatemala: 2 Total: Neither happened 0 times 1 only happened 2 times 2 only happened 5 times 1 and 2 happened 5 times 1 with or without 2 happened 7 times 2 with or without 1 happened 10 times I've now done two league matchdays, one Champions League matchday for half the clubs in the Group Stage, one World Cup knockout round, and one double round-robin World Cup qualifying group for a total of 46 games. Total: Neither happened 1 time 1 only happened 13 times 2 only happened 14 times 1 and 2 happened 18 times 1 with or without 2 happened 31 times 2 with or without 1 happened 32 times This sample supports my hypothesis that I expected 2 to happen slightly more than 1.
Good work and interesting stuff. Why do you hypothesize that 2 is more likely than 1 to be the final score? Any specific reasons? A big counter example that springs to mind is games between a big team and a minnow. Many a times the game is well settled in the first half and the second half is goal less and/or the big team takes off their key players. At a 90% level of significance, testing for significant differences in the proportions, here is what we see: EPL Nov 20-22 total games = 10 total occurrences of '1' = 8 total occurrences of '2' = 6 * note small sample size Not significantly different. ---------------------------------------------- UEFA matchday 5 total games = 8 total occurrences of '1' = 3 total occurrences of '2' = 7 * note small sample size Significantly different at 90% LoS --------------------------------------------------------------- total sample together that you've chosen total games = 46 total occurrences of '1' = 31 total occurrences of '2' = 32 Not significantly different. We need more ( and larger samples) before judging if 2 is more prevalent than 1. My guess is that they wouldn't vary much.
I already said "because the second half has more stoppage time, with a large sample I would expect 2 to happen a little more often than 1." Also I was comparing each half to the "final result" (ignoring the margin of victory in games with winners), not the "final score." If I compared to the "final score" 1 and 2 could never both happen because the final score is the sum of the halves. Another thing I could look at is in games with winners what percent of the time the winning team outscored the opponent in both halves, in one half with the other half score equal, and with each team outscoring the other in one half. For draws with goals I could look at what percent of the time the score was even at halftime and what percent of the time each club outscored the other in one half.