Electoral College contest underway!
With two weeks and a day until the election, it's time to get the second quadrennial BrendanLoy.com Electoral College Contest started. You can enter the contest here.
The deadline to enter is noon EDT on the day before the election, Monday, November 3. If you enter and then change your mind, you can re-submit your picks at any time up until the deadline, either by filling out the above-linked form again from scratch, or by simply e-mailing me your new "embed code" and "link URL." (You'll understand what that means after you fill out the form.) Your latest entry will automatically replace any earlier entries.
If you have any technical difficulties signing up, particularly with the Flash map, or if anything is confusing, please let me know. Thanks!
As I explained previously, I've simplified the rules this year: the winner will simply be the person who gets the least number of electoral votes wrong. The "differential" between the candidates' EV totals and each contestant's prediction of their EV totals will be used only as a tiebreaker, not as a determining factor in the standings, as it was four years ago. This mostly eliminates the advantage of making multiple "offsetting" errors. For a further elucidation of the significance of this change, go here.
There is no requirement that the contest winner must necessarily have predicted the correct overall winner in the election. Fundamentally, this is a state-by-state prediction contest, not a national prediction contest. However, there is a tiebreaker that rewards getting the overall winner right. More on this in the "UPDATE" below.
Just like four years ago, I will abide by the final popular-vote result in each state (and, in the case of Maine and Nebraska, each congressional district), as certified by the relevant governing authority in that state; “faithless electors” will not be taken into account. A winner will be declared as soon as the state-by-state election results are sufficiently complete that such a declaration is possible.
Initial results will be posted on Election Night, during the planned Mother of All Liveblogs, using my nifty map / table / liveblog / livechat interface. :) However, those initial results are subject to change in the event of reversed media “calls,” recounts, lawsuits and so forth. I will have the ultimate discretion to determine, within the letter and spirit of the rules and in the interests of fairness, how to deal with unforeseen controversies that may arise in the event of disputed state elections.
The main prize is bragging rights, but the winner will also get a fun CafePress goodie of some kind. Four years ago, Mike Wiser's perfect map earned him a coffee mug with that selfsame map printed on it, along with a printed message declaring him the contest "champion":
Anyway, good luck, everyone! Again, the entry form is here.
UPDATE: I've changed the tiebreakers around a bit, and I'm moving their full text into this post:
Tiebreaker #1: Whoever comes closer, in absolute value terms, to predicting Obama's electoral-vote margin of victory/defeat, wins. (This is the "differential" rule.)
Tiebreaker #2: Any contestant who predicts the correct national EV winner finishes ahead of a tied contestant who predicted the wrong winner. For purposes of this rule, a 269-269 map is regarded as predicting an Obama victory. However, if the actual result is a 269-269 tie, this rule is inoperative, and we skip to tiebreaker #3.
Tiebreaker #3: Whoever has the least number of states wrong, wins. (The District of Columbia counts as a "state." As for Maine and Nebraska, if a contestant gets the at-large result right, but one or two congressional districts wrong, this will be counted as a "half-state" wrong.)
Tiebreaker #4: Whoever comes closer, in absolute value terms, to predicting Obama's national popular-vote margin of victory/defeat, wins.
Tiebreaker #5: Any contestant who correctly predicted the state with the closest popular-vote margin, in percentage terms, defeats any tied contestant who failed to do so.
Tiebreaker #6: Whoever comes closer, in absolute value terms, to predicting the number of Democratic Senate seats, wins.
Tiebreaker #7: Whoever comes closer, in absolute value terms, to predicting the number of Democratic House seats, wins.
Tiebreaker #8: Whoever comes closer, in absolute value terms, to predicting Obama's raw vote total in Colorado, wins.
In all cases, "faithless electors" will not be factored in, nor will any post-election party switches in Congress.
The reason I added an explicit "correct national EV winner" tiebreaker is because I realized I misspoke earlier when I said the "differential" tiebreaker guarantees that a tie would go to the contestant who got the right winner. For instance, suppose Obama wins 273-265, based on this map. Contestant A has a perfect map except he thought Obama would win Nevada (5 EVs). Contestant B has a perfect map except he thought McCain would win New Mexico (5 EVs). So, both contestants made 5 EVs worth of mistakes, and both have a "differential" of 10. (Contestant A predicted an Obama margin of +18; Contestant B predicted an Obama margin of -2; Obama's actual margin was +8, halfway in between the two.) In my view, Contestant A should win in this scenario, because he got the actual winner right.
UPDATE 2: Upon further review, I've decided to have "differential" be the first tiebreaker, and "right winner" the second tiebreaker. I could make a case for either order, but I think, on balance, this order is fairer. Of course, in the scenario I described above, this would make no difference, since the differentials are identical, but in some cases it could make a difference. So, the way the rules are now written, the "right winner" rule only applies if the tied contestants have the same number of EVs wrong and the same EV "differential."
The deadline to enter is noon EDT on the day before the election, Monday, November 3. If you enter and then change your mind, you can re-submit your picks at any time up until the deadline, either by filling out the above-linked form again from scratch, or by simply e-mailing me your new "embed code" and "link URL." (You'll understand what that means after you fill out the form.) Your latest entry will automatically replace any earlier entries.
If you have any technical difficulties signing up, particularly with the Flash map, or if anything is confusing, please let me know. Thanks!
As I explained previously, I've simplified the rules this year: the winner will simply be the person who gets the least number of electoral votes wrong. The "differential" between the candidates' EV totals and each contestant's prediction of their EV totals will be used only as a tiebreaker, not as a determining factor in the standings, as it was four years ago. This mostly eliminates the advantage of making multiple "offsetting" errors. For a further elucidation of the significance of this change, go here.
There is no requirement that the contest winner must necessarily have predicted the correct overall winner in the election. Fundamentally, this is a state-by-state prediction contest, not a national prediction contest. However, there is a tiebreaker that rewards getting the overall winner right. More on this in the "UPDATE" below.
Just like four years ago, I will abide by the final popular-vote result in each state (and, in the case of Maine and Nebraska, each congressional district), as certified by the relevant governing authority in that state; “faithless electors” will not be taken into account. A winner will be declared as soon as the state-by-state election results are sufficiently complete that such a declaration is possible.
Initial results will be posted on Election Night, during the planned Mother of All Liveblogs, using my nifty map / table / liveblog / livechat interface. :) However, those initial results are subject to change in the event of reversed media “calls,” recounts, lawsuits and so forth. I will have the ultimate discretion to determine, within the letter and spirit of the rules and in the interests of fairness, how to deal with unforeseen controversies that may arise in the event of disputed state elections.
The main prize is bragging rights, but the winner will also get a fun CafePress goodie of some kind. Four years ago, Mike Wiser's perfect map earned him a coffee mug with that selfsame map printed on it, along with a printed message declaring him the contest "champion":
Anyway, good luck, everyone! Again, the entry form is here.
UPDATE: I've changed the tiebreakers around a bit, and I'm moving their full text into this post:
Tiebreaker #1: Whoever comes closer, in absolute value terms, to predicting Obama's electoral-vote margin of victory/defeat, wins. (This is the "differential" rule.)
Tiebreaker #2: Any contestant who predicts the correct national EV winner finishes ahead of a tied contestant who predicted the wrong winner. For purposes of this rule, a 269-269 map is regarded as predicting an Obama victory. However, if the actual result is a 269-269 tie, this rule is inoperative, and we skip to tiebreaker #3.
Tiebreaker #3: Whoever has the least number of states wrong, wins. (The District of Columbia counts as a "state." As for Maine and Nebraska, if a contestant gets the at-large result right, but one or two congressional districts wrong, this will be counted as a "half-state" wrong.)
Tiebreaker #4: Whoever comes closer, in absolute value terms, to predicting Obama's national popular-vote margin of victory/defeat, wins.
Tiebreaker #5: Any contestant who correctly predicted the state with the closest popular-vote margin, in percentage terms, defeats any tied contestant who failed to do so.
Tiebreaker #6: Whoever comes closer, in absolute value terms, to predicting the number of Democratic Senate seats, wins.
Tiebreaker #7: Whoever comes closer, in absolute value terms, to predicting the number of Democratic House seats, wins.
Tiebreaker #8: Whoever comes closer, in absolute value terms, to predicting Obama's raw vote total in Colorado, wins.
In all cases, "faithless electors" will not be factored in, nor will any post-election party switches in Congress.
The reason I added an explicit "correct national EV winner" tiebreaker is because I realized I misspoke earlier when I said the "differential" tiebreaker guarantees that a tie would go to the contestant who got the right winner. For instance, suppose Obama wins 273-265, based on this map. Contestant A has a perfect map except he thought Obama would win Nevada (5 EVs). Contestant B has a perfect map except he thought McCain would win New Mexico (5 EVs). So, both contestants made 5 EVs worth of mistakes, and both have a "differential" of 10. (Contestant A predicted an Obama margin of +18; Contestant B predicted an Obama margin of -2; Obama's actual margin was +8, halfway in between the two.) In my view, Contestant A should win in this scenario, because he got the actual winner right.
UPDATE 2: Upon further review, I've decided to have "differential" be the first tiebreaker, and "right winner" the second tiebreaker. I could make a case for either order, but I think, on balance, this order is fairer. Of course, in the scenario I described above, this would make no difference, since the differentials are identical, but in some cases it could make a difference. So, the way the rules are now written, the "right winner" rule only applies if the tied contestants have the same number of EVs wrong and the same EV "differential."
posted by Brendan Loy at
10/20/2008 12:59:00 PM
