UPDATE (3:45 PM): Note that the success rates cited below do not account for the dozens and dozens Coleman challenges that have gone into the "blue folder" -- these are challenges not to the marks on the ballot but "process" challenges related to its propriety (such as whether the ballot's duplicate is missing). If we count the blue folder ballots as unsuccessful challenges, then Coleman's success rate is more like 7% -- however, it would be premature to do that since we simply don't know enough about them.
UPDATE (3:30 PM): Coleman was having a really terrible run of things for an hour or so there after lunch, with something like 5 or 6 challenges upheld out of more than 100 ballots evaluated. Most of these ballots were from Minneapolis and other Democrat-friendly places in Hennepin County, so this is perhaps no surprise. Overall, his success rate is now closer to ~10-11% than ~12-13%.
Also, please do not take any of the numbers you see here as gospel. Everyone is working off some or another kind of best guestimate.
___
With the Canvassing Board today having begun ruling on ballot challenges made by the Coleman campaign, we may now have a slightly better idea of where we stand in Minnesota's recount process.
Through lunchtime today, by my highly unofficial count, Coleman has won 33 of his 265 challenges, or 12.5%. By contrast, on Tuesday and Wednesday, Al Franken won 115 of his 393 challenges, or 29.4% -- again, by my highly unofficial count. (That Franken's success rate is much higher is not surprising since he had significantly fewer challenges outstanding)
The guys at The Uptake reported earlier today that Coleman has 1,021 challenges presently outstanding to the Canvassing Board. If he continues to win challenges at his current rate, he'd wind up with 127 successful challenges.
Franken apparently has some challenges that have yet to be resolved too. These are so-called "blue folder" ballots that were challenged for process (e.g. the absence of a duplicate ballot) rather than a voter intent reasons. There are something like 20-30 Franken ballots in the blue folder, which at a 29% success rate would net to another 7 ballots or so, bringing his total to 122 successful challenges.
If Coleman wins 127 challenges and Franken wins 122, that would mean a net of 5 votes for Coleman. If the Franken campaign's estimate was correct that it was winning the re-count by 4 votes assuming that all challenges would be rejected, that would produce a Coleman win by a grand total of ... one vote!
However, there remain any number of contingencies that make these estimates highly imprecise:
1. There were some challenges made by each candidate that apparently ended in the opponent's pile ... some Coleman challenges, for instance, in Franken's stack of ballots. I am confused by these and have tried to avoid counting them for now, but they could certainly make a difference of 5-10 ballots at the margins.
2. Coleman in particular has had a lot of "blue folder" ballots, which will not be addressed until after Christmas. Although my hunch is that very few blue folder challenges are liable to be successful (they probably point toward issues that are outside of the Canvassing Board's jurisdiction), it is hard to know for certain.
3. Both campaigns are now talking about adding back challenges that they had previously withdrawn.
4. We don't know if the Coleman challenges counted so far are representative of his efforts as a whole. (For what it's worth, the Board has not yet addressed any Coleman challenges from Minneapolis, St. Paul, or Duluth, which are highly blue cities that may lower Coleman's success rate).
5. There is no way to independently verify the Franken campaign's estimate that it was ahead by 4 ballots assuming that all challenges would be rejected.
6. We have not accounted for "fifth pile" absentee ballots that, if and when they are counted, could produce a net gain of as many as 100 votes (or more) for Franken.
7. Pretty much everything that has happened so far, and everything that happens from here on out, can and quite possibly will be the subject of a court challenge.
But other than that, we've got it all figured out...
12.18.2008
Minnesota Lunchtime Update: It's Gonna Be Close
by Nate Silver @ 3:21 PM
Subscribe to:
Post Comments (Atom)

58 comments
I think I called this election, Franken by 7 votes weeks ago.
Do I get a cookie if I'm close?
Amazingly close.
I think they should hold a special election to figure this out, but I guess there is no basis in law for this common sense fix.
Fred -- do you mean a runoff? Well, as long as we're wishing for things that aren't in Minnesota law, instant-runoff voting would be even better, since it would be a vote of the full electorate from Election Day, rather than the small fraction who will come out for a runoff or special election.
wv: derism -- a religion based on mockery
The city of Minneapolis, MN is taking up instant run off voting-we overwhelmingly voted for it, but lawyers for the opposing side have taken it to court. Hopefully we'll have state wide instant run off voting soon! Hopefully against Pawlenty when he runs for Republican nomination in '12
I'm still putting my faith in the 27 vote victory earlier posted.
LDoF: You get a cookie if Franken wins by any amount.
If Coleman wins, as they say, "no biscquit".
So if he legally changes his name to Al "Lizard People" Franken, does he get the vote?
The C challenges in the blue counties might also tend to be more successful than in red counties, in case there is any local bias which the Board tends to correct.
Still, this one is clearly tied to within the error bars.
It sounds likely that the absentees will get in, but I'm not confident that we know how they'll go.
I have to admit that I have more or less stopped looking at the details of this senate race. I find comments like "can Franken pull ahead?" and such, rather meaningless because election day was long ago and actually it's all decided... There must be a different solution than letting the courts decide the outcome.
But I lately I've had the suspicion that Coleman is about to fall very deep. He's going to lose the race, might have problems with the law and he might end up in debt if he keeps on using campaign money for legal action.
I'm surprised that the Coleman is trading so low on intrade (28.5). Is this just a reaction to the "fifth pile" or is there something else going on here?
All of this for an election in which even a PERFECT count wouldn't tell us which candidate is the voters' choice because neither got anywhere near 50%. What's the point?
Nate, I think we need to start the push for "instant runoff" voting. Though it doesn't necessarily avoid situations like this wherein the totals are super close, at least the effort we'd be putting in would have a chance at determining which candidate the electorate actually prefers. This would also avoid problems like Georgia had in making people turn out to vote in a race they already voted for.
Think about it.
***
THIS IS GREAT NEWS FOR McCAIN!
All-
(IMHO)-the Strib live stream is better than the UPtake live stream because the strib shows each ballot and Uptake does not.
wv-gulec where Cheney should land for specific war crime and for being evil.
everyone is fucking coughing.
I'm surprised that the Coleman is trading so low on intrade (28.5). Is this just a reaction to the "fifth pile" or is there something else going on here?
Probably because the smart money feels that when the issue ends up in the Senate the Democratic majority will side with Big Al.
Re Instant Runoff:a good idea but can you imagine doing a hand recount on those ballots ?
redshift, ron, opus,
Many many posts ago, Minnesota Latte Liberal and I had a long discussion about IRV (starting here). I am against IRV because it violates some pretty fundamental voting system criteria, and any advantages it offers are also offered by many other voting systems.
Although I should point out that Redshift is absolutely right that IRV is preferable, both from a theoretical and practical standpoint, to the two stage runoff system used in Georgia's senate race.
Nate - thanks for all your work on this. You've been the best source of information as to what the current totals actually mean to the eventual result, or at least what they are likely to mean.
I too think Franken will win by a few votes - I'll say 34.
Each time the Chair moves to
award a vote, he says,
"All in favor, say Aye."
"Opposed, ???"
What is he saying after "Opposed"?
Thanks.
@ Seth Teller:
He's saying, "Opposed, same sign" which apparently means "if you are opposed, say the same thing I asked you to say if you were in favor, i.e., say 'Aye' again."
I don't know why they do it this way. I prefer "Aye" vs. "Nay", but maybe in a large group, it's hard to compare the relative volume of different "signs."
Any way, everyone votes "Aye" no matter what, it's just when they say it that counts.
WV - aysingum - An alternative vocal way of voting.
Hey Nate, if this is anywhere close to Franken +27, somebody's going to say your estimate several weeks ago was just dumb luck. But I would just say your approximation was a very good one.
wv: untru (true dat)
Seth Teller_
"Opposed- same sign"
Which means say Aye if you oppose the motion
I hope Franken wins, but no matter who wins and under what circumstances, the opposing side is going to claim fraud.
It's amazing how huge a store of dumb luck Nate has hidden away.
off topic- good news.
VA-05 is now official. We're done with Goode.
-> total House pickup of 21, and that includes the welcome surprise loss of Jefferson in LA.
The Star Tribune's live feed has it to Coleman +5. Anyone think Franken will be ahead by the end of the night?
@ Efrem:
No. They're done for the night
Blogger Michael (mbw) said...
Has Goode conceded?
No, because they just stopped counting for the day.
My question is how many "un-confusing" (i.e. blue book, green book, etc) Coleman challenges remain to counted? And when will the supposedly withdrawn challenges from both campaigns going to be truly added to the counts.
Are both campaigns just playing possum with those "withdrawn but still not counted" challenges?
under IRV it is possible to lose by getting too many votes.
@ SHMA
The above is from one of your November 23 posts.Can you give me an example of how this can happen?
@hill.tops "Has Goode conceded?"
Who cares? The election recount is now certified by a court.
These are strange numbers:
Coleman has picked up 253 votes, Franken 432. But, the sum of the Coleman votes from Franken+Coleman challenges is 249 (only 4 less than 253), while Franken's number is 319 (over 100 less than 432).
Is Franken picking up votes from ballots Coleman withdrew on the spot that aren't being counted as Coleman challenges?
Blogger Michael (mbw) said...
cuz numbnutz can file an appeal.
so either it stops today, or goes on for a few more weeks.
The Minnesota Supreme Court just ruled that the wrongly rejected absentee ballots should be counted. They emphasized they should be evaluated with a uniform standard but I think the canvassing board has already shown they are capable of setting a valid standard and sticking to it.
Coleman can appeal to the US Supreme Court but court experts regard that as unlikely to be successful.
Thanks joseph,you made my night.
I don't see how Coleman wins with only a 10% win rate. I thought he reduced his challenges to 800 (used to be 1000, subtracted 400, and added 200).
That will leave him about 30 to 40 votes short (assuming Franken's campaign was correct about the 4 vote margin if all ballots were rejected).
It would be painful if Franken lost only because the absentee ballots were counted.
The idea of "instant run-off" voting sounds good, but I am worried about the fact that it is very different from what voters are used to and would force them to consider things they haven't had to. How many voters will (or won't) understand it and utilize it?
I know that the city of Cambridge, MA had a ssytem of sequential preference voting for city council at one point. (I don't know if they still do.) There were 9 at large seats on the city council and voters could rank order up to 9 candidates. A certain percentage of votes guarranteed a seat. The lowest candidate was eliminated and the votes were then awarded on the second count to those voters' second choice. Those candiates who achieved the threshhold vote were elected on the first count, and some proportional method of assigning "excess votes" to the second choice candidates on the seond oount would be used. The process get repeated for each round of counting until 9 candidates were elected. As I recall, the one year I was living there when this happened the final count wasn't arrived at until Friday froma Tueday election.
I know that not all voters used all nine votes. I would imagine that there was some scholarly research on those voting patterns, probably by researchers from MIT as Hasrdard's government department was strongly anti-behavioralist at that time.
I recall immediately after a senior honors oral exam that the head of the Government department asked me what I was going to be doing the next year, and when I told him that I was going to go to Michigan for graduate work a sneer crept into his expression and he said something like "Oh, so you're gong to play with comoputers."
(Barney Frank, at that time a grad student and a teaching fellow for the chair, was a member of the examining grooup.)
Two things:
#1 I am a fan of Instant Runoff
#2 I am less of a fan than I use to be.
Recently I learned of other possible reforms besides IRV. A relatively few number of people have evaluated these options and opined. They have pretty polarized views and it seems to me each side has some valid arguments and each side has some, more ahhh, seemingly manure- type arguments.
With the greatest of respect to the folks at Fairvote.org and RangeVoting.org, something as important as our democracy ought not be debated by a few theorists and 501(c)3 activists. Every community thinking about reform should be aware of the options, have its *own* conversation about them, and contribute its thinking to the national conversation as a whole.
All that said, I'm still an IRV guy, for now. However, I do think communities should take a hard look at at the various alternatives and then consciously decide which they think is the best one to try in their own town. Maybe its IRV, but maybe its one of the others.
Steve E
I like IRV. Always have.
But can you imagine the kinds of trouble some nitwit like Mr/Ms Lizard People would have with an IRV ballot? He/she would probably mark EVERY candidate as a first preference and then put "Lizard People" as 0.5.
There are times when I think some people just shouldn't vote at all.
The MN Supremes did NOT order the counties to count the erroneously rejected absentee ballots. They said count ONLY the ones that BOTH campaigns agree were erroneously rejected. That'll never happen. They WIMPED out, just read the great Justice Alan Page's scathing opinion in dissent.
Looks like Franken may have to take this to federal court, unless, highly unlikely, Coleman concedes after the Canvassing Board certifies the election in favor of Franken (likely).
So, Coleman appears to up by 2 votes.
A few questions:
1. How many more Coleman challenges will be decided tomorrow?
2. Have all of the Franken challenges been decided?
3. How many Coleman and Franken challenges were withdrawn and when will they be added to the totals?
4. Do we really have reason to believe that Franken will get more of the absentee ballots when they are counted?
5. What with the so-called "duplicate" votes that Coleman keeps claiming exist?
6. Is there another category of votes I should know about?
Thanks!
Can't wait for KWR to parse today's results. I missed some of the proceedings as I was driving to the CB meeting, where my laptop battery eventually died.
@shma :)
We did, and I would only restate that I am still a fan of IRV with a proviso that the IRV kicks in under specially defined set of rules, not just during lack of majority.
@green libertarian,
just got home, have not read the ruling yet, but from the MPR and Air America and telephone analysis :)) I get the interpretation that what all counties and campaigns have to agree on is a uniform standard under which to count the wrongly rejected votes, not agree on every vote. Everyone I heard from might be wrong, and I do not discount that, but I would not lose hope quite yet.
I had Allan Page's dissenting opinion read to me as I drove, and I tend to concur with him up to a point regarding chaos. Mob rule should not be, the SOS ought to take helm at this point and proclaim: "these are my standards according to the MN law. Deal with it, peons!" Be firm, lay down the lay for the counties now that he's got the umbrella SC ruling.
~ Latte
wv: tingla. I've got this happy tingla feeling after watching so many Coleman challenges get rejected today live.
I tend to believe that the Star Tribune's internet vote on the ballots will be 99.9% accurate for withdrawn ballots[they were withdrawn because they were frivolous so anyone voting should be able to figure them out. Looking at the resolved challenges, ballot box stuffing doesn't appear to be a problem]. From that it seems that Franken +89 is the most accurate current state of the race that we have with the biggest question mark in my mind being how the star tribune is allocating the green/blue folder ballots.
@ William Land.
I am going to try :)
1. How many more Coleman challenges will be decided tomorrow?
afair we were about half way done today. if we subtract the number (~150 or so duplicates in the blue folder, your ?5) to be ruled on first thing tomorrow morning at 9AM, we get a fairly reasonable number under 400. Very roughly.
2. Have all of the Franken challenges been decided?
Franken's got another 200 or so they could introduce pending the Board's ruling on #5 per above. They are unlikely to add more imho because a) things are breaking their way anyway, b) that will only delay the proceedings and time works against them and c)the Board, based on what I saw today after 5PM in the live discussion with both Trimble and Elias will NOT approve Coleman's motion on this. They will not uphold challenges on duplicates as double votes if only because of the equal protection issue (Franken was not allowed to challenge in Wright County on this basis per local judge who happened to be SoS Office Rep).
3. How many Coleman and Franken challenges were withdrawn and when will they be added to the totals?
Coleman started with a number north of 3200 (look at the old 538 topics and Strib archives on the matter). Now he is down to 400 left. Franken was down to 390 as per Nate's estimate yesterday out of 115 less challenges than Coleman.
Afair, the w/d's will be added to the totals by the SoS when they tabulate all the results over the weekend. I honestly do not expect anything quasi-official from Ritchie before Monday.
4. Do we really have reason to believe that Franken will get more of the absentee ballots when they are counted?
People in the know here and on pollster.com keep referring to the early voting (which would be absentee only in MN) polls that seemed to indicate that as the election date drew close there were 1-2 polls showing that folks who had already voted (absentee) broke +8 for Franken. Again, no link, going from my memory alone.
5. What with the so-called "duplicate" votes that Coleman keeps claiming exist?
This is fascinating and has been discussed ad nauseam in earlier threads. Basically, some absentee ballots are printed off wrong format oversees/filed electronically, etc. and a duplicate has to be made for each. Then each original will bear a mark "original #N" and the corresponding duplicate that was fed through the counter ought to have the same "duplicate N" marking. Often [enough] times, for whatever reason, the dupes cannot be found during the recount. Hence [false] concern by Coleman/Trimble about double voting, now handily debunked by Elias. And that, folks, was a thing of beauty. A beauty, I tells ya!
6. Is there another category of votes I should know about?
Provisional ballots. Those cast on election night where the credentials of the voter had yet to be confirmed. Sorry about this can of worms :)
~Latte
wv crock
Latte,
Do you know if the 477 Coleman challenges reported by the STrib as reviewed today included those which Coleman withdrew at the last minute (400 of them)? I thought some were brought up and withdrawn during the count? Or were all of those taken out before the counting started?
MN Latte Liberal:
Thanks!!!
This is incredibly complicated.
I knew nothing about the provisional ballots...any idea how many there are?
@just_looking,
that's a great question, and I do not know their fate. I strongly suspect that they were left in the folders of copied challenges that all 5 board members got just because there was not time to withdraw them physically. And when those ballots came up in the folder, Trimble would announce "withdrawn", and the Board would instantly vote to assign the vote to Franken (or 3rd party as the case may be). However, there is no way for me (or anyone but a campaign insider) to know if those spot-withdrawals are all part of the 400 or some were additional newer withdrawals yet or if a part of the 400 was removed prior to the folder (which as I mentioned, I strongly suspect was the case).
In Franken's challenges case the situation was much the same, and it caught a lot of ppl by surprise when they actually got within a handful of ballots yesterday and then finished. At least that was my read.
@ William Land,
I have no idea as to the fate of the provisionals, and have been trying to figure/find that out. I was going to ask Noah Kunin (of theuptake.org) when I saw him in person during the hearings, but he was incredibly busy, and I was just a random dude off the street to him.
During the recount, none of the observers or our site leader knew the fate of the provisional ballots, and I never got a chance to ask the county election head.
So, they remain a great mystery to me. I want to hope they were already assigned, but am unsure as to when that would've taken place.
~ Latte (off to bed)
wv ingwp: the great northern vowel drought.
Thanks again, Latte---and great rundown of the variable factors. Rather than "parse today's results", I instead modified my pre-today prediction in earlier thread---with the knowledge that Franken had won 57 vote-challenges on Tue. & Wed., rather than 33 as I thought based on that thread. The +24 improvement looks significant, but actually it made little difference, because I was basing more on concrete reports and well-founded prior assumptions about the vote-fertility of the ground and the campaigns' behavior, not just extrapolating projections. And algebraically somethings canceled out. My concrete predictions for tomorrow:
() When Yellow Medicine County is reached---assuming the "Coleman dupe challenges" are summarily ruled out and before any "200 extra challenges" from either candidate are considered---the running tally at TheUptake (and the STrib if in-sync) will show "Coleman +270". That's unchanged from what I indicated 24 hours ago.
() At the same time, "Other" will hit 282. And by my figuring, that represents overall Franken +8, having gained a staggering +4 in the challenge phase! (Not the upwards-of-20 gain I thought, because vote-challenges were notably more frequently granted than I expected, a factor favoring Coleman despite Franken's initial success.)
Today, "Coleman" went from +238 to +253, and "Other" went from 114 to 198. Based on the % of Coleman's list gone thru, the numbers instead blandly project to hitting about +265 and 270, respectively. So that indeed would be a Franken +21 gain and +25 overall margin from this phase (ceteris paribus), hence my crack there about Nate's "+27". However, today's afternoon session was dominated by Minn, so I will hold to my original numbers. (Well, I could say the lower totals are still within my error bars, but as a chessplayer I'm temperamentally avid to hit my evaluations exactly.)
Franken will at least try to re-introduce 100-200 challenges. Indeed if my "8%-er threshold" is right, then he could expect to net 6--10 votes from that, and with MN SC definitely causing a delay in when he'd expect the improperly-rejected absentees to be counted, the present value of that potential handful of votes might be considerably more than he reckoned on at the start of the week.
Of course, as Nate is even saying, everyone is working from guesstimates. The one difference I claim is that as much as possible, I'm basing on:
() reports---by M. Elias, "mediapost", and "UWBHS";
() prior assumptions: more "vulnerable" ballots are from Dems, but given equal populations of Dem and Rep vulnerables, the results will average out the same; and
() reasonable inferences based on the campaigns' behaviors to date.
I munged part of my post by a last-minute mouse-edit: before "Franken will at least try" was "() I also predict that ", and "Minn." was supposed to be short for Minneapolis :-).
And they may be 5% further along in Coleman's list than I thought: The STrib here shows
Olmstead County | ROCHESTER W2 P2 | Challenge 27
as "Ballot 1 of 470", but the next ballot is same-county, Challenge 26. Hence I infer they are in last-in/first-out order, i.e. that the above was the last ballot reviewed today. This means they finished on ballot 634 of Coleman's Excel list of 1,016. Assuming the withdrawals and "blue" are evenly distributed, the progress is 634/1016 = 62.4%. That projects Coleman's 15 today into 24, putting him on +262 from +238, and projects "Other" from +84 today into +134, leaving it at 249 which is definitely below my error bars. Basically it represents Coleman's "excess" challenges netting zilcho, i.e. he projects to knock out no more initial-Franken votes than Franken did vice-versa with far fewer such challenges. Or it may represent my simply being off by 30+ there. I guess I can't say more than "we'll see".
If Minnesota had used Instant runoff, there might still have been a recount, and it would have been much harder to discern voter intent with voters being asked to make multiple votes for each contest.
Worse, there would be argument and debate as to whether the votes were "transferred" accurately, since IRV requires a complex formula to tabulate the votes.
North Carolina tested IRV in a small city, and the best election officials in the state (supposedly) couldn't even count 3,000 votes accurately, and the votes had to be recounted - all of them, and all 3 rounds.
In Australia, where they have a form of IRV, it can take anywhere from 1 week to a month to get the results of a single contest election.
With IRV, if you had to do a recount, it would be worse than recounting multiple elections, it would be a nightmare.
If Minnesota voters can't mark a simple ballot correctly, imagine what they would do when asked to make multiple choices for each contest.
Crazy.
+135 I meant---hit wrong key.
WV "entyping" (no kidding!): Misteaks happen entyping.
If "Other" stays closer to 250 than 280, while "+Coleman" hits the mid-260s, this will mean that Coleman won no more vote-challenges than Franken did.
My above-linked analysis implicitly presupposes that (a) having more "vulnerable Dems" will translate into more spoiled Franken votes.
I wonder if experience with recounts shows instead that (b) the total populations of winnable vote-challenges are roughly equal.
If (b) is the rule---as "Other" near 250 will support---then only the difference in the # (and quality) of winnable non-vote challenges matters, and that would be a "prior explanation" for Franken's "real lead" marching from +4 to somewhere around 30.
Indeed, in the SoS spreadsheet of Franken's challenges (which may include about 20 Coleman interlopers or doubly-challenged ballots?):
18 are resolved as "NO ONE/IDENTIFYING MARK, and
27 are resolved as "NO ONE/OVERVOTE".
These two categories strike me as independent of "voter vulnerability", and (hence?) as a Dem/Rep wash. Insofar as they may comprise a lot of Franken's said-to-be-57 upheld challenges of Coleman votes, this seems to be evidence for (b).
In that case, I should have projected 29 Coleman wins from non-vote challenges, + 2 not 3 "flips, + the same 57 vote-challenge wins as Franken got, for 88. Which is 33 short of the 121 that would have produced a tie.
One other factor: I haven't checked whether the "Coleman interlopers" on Franken's list also appeared on Coleman's list. If so, then some portion of the Coleman +7 and Other +6 resulting from challenges the STrib listed as in Coleman's 20 should be counted as part of my prediction. (The spreadsheet formats are different enough so I can't simply do a merge and unique-lines sort.)
Absenting that last factor, it looks like Coleman is coming in 3-5 short of my 29, and from the mid-morning report on TheUptake's live blog that he had knocked out 44 initially-Franken votes, maybe 3-5 short of Franken's 57 too. That would bring this up exactly to Nate's 40-vote margin, which I saw while in the middle of gathering the above figures.
Opus,
Not sure if you're still checking this thread, but you can find the simplest example (a 3 candidate situation) at the wikipedia page for monotonicity. Again, this is the simplest example, not the only one. Nor are these the only numbers for which this example works. You can play around with the numbers and see that you get the same violation with 34/35/31 instead of 39/35/26. You can also get another perverse result by switching the final and initial vote tallies. Starting with 49/35/36 and letting 10 voters abandon Andrea for their second choice (say she called them a bunch of racists, or something) causes Andrea to win. In other words, despite these voters consciously trying to make Andrea lose, they have instead given her a victory.
情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,潤滑液,自慰套,威而柔,威而柔,威而柔,威而柔,自慰套,跳蛋,按摩棒,充氣娃娃,自慰套,潤滑液,威而柔,FleshLight,跳蛋,按摩棒,充氣娃娃,跳蛋,按摩棒,FleshLight,充氣娃娃,情趣商品,情趣網站,情趣網站,潤滑液,性感內衣,充氣娃娃,按摩棒,情趣精品,跳蛋,情趣網站,情趣商品,跳蛋,FleshLight,充氣娃娃,情趣內衣,情趣精品,按摩棒,威而柔,自慰套,成人玩具,Nexus,lelo,聰明球,後庭,後庭g點,g點,美國fleshlight,STU訓練大師,Fleshgirls,Toys Heart,Tenga,日本 Vibratex,日本Toys Heart ,日本Tenga,美國aneros,rudeboy,英國rudeboy,英國Rocksoff,德國Fun Factory,Fun Factory,英國甜筒造型按摩座,甜筒造型按摩座,英國Rock Chic ,瑞典 Lelo ,英國Emotional Bliss,英國 E.B,荷蘭 Natural Contours,荷蘭 N C,美國 OhMiBod,美國 OMB,Naughti Nano ,音樂按摩棒,ipod按摩棒,美國 The Screaming O,美國TSO,美國TOPCO,美國Doc Johnson,美國CA Exotic,美國CEN,美國Nasstoy,美國Tonguejoy,英國Je Joue,美國Pipe Dream,美國California Exotic,美國NassToys,美國Vibropod,美國Penthouse,仿真按摩棒,矽膠按摩棒,猛男倒模,真人倒模,仿真倒模,PJUR,Zestra,適趣液,穿戴套具,日本NPG,雙頭龍,FANCARNAL,日本NIPPORI,日本GEL,日本Aqua Style,美國WET,費洛蒙,費洛蒙香水,仿真名器,av女優,打炮,做愛,性愛,口交,吹喇叭,肛交,魔女訓練大師,無線跳蛋,有線跳蛋,震動棒,震動保險套,震動套,TOY-情趣用品,情趣用品網,情趣購物網,成人用品網,情趣用品討論,成人購物網,鎖精套,鎖精環,持久環,持久套,拉珠,逼真按摩棒,名器,超名器,逼真老二,電動自慰,自慰,打手槍,仿真女郎,SM道具,SM,性感內褲,仿真按摩棒,pornograph,hunter系列,h動畫,成人動畫,成人卡通,情色動畫,情色卡通,色情動畫,色情卡通,無修正,禁斷,人妻,極悪調教,姦淫,近親相姦,顏射,盜攝,偷拍,本土自拍,素人自拍,公園露出,街道露出,野外露出,誘姦,迷姦,輪姦,凌辱,痴漢,痴女,素人娘,中出,巨乳,調教,潮吹,av,a片,成人影片,成人影音,線上影片,成人光碟,成人無碼,成人dvd,情色影音,情色影片,情色dvd,情色光碟,航空版,薄碼,色情dvd,色情影音,色情光碟,線上A片,免費A片,A片下載,成人電影,色情電影,TOKYO HOT,SKY ANGEL,一本道,SOD,S1,ALICE JAPAN,皇冠系列,老虎系列,東京熱,亞熱,武士系列,新潮館,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,整型,水噹噹,貸款,貸款,信用貸款,宜蘭民宿,花蓮民宿,未婚聯誼,網路購物,珠海,下川島,常平,珠海,澳門機票,香港機票,婚友,婚友社,未婚聯誼,交友,婚友,婚友社,單身聯誼,未婚聯誼,未婚聯誼,婚友社,婚友,婚友社,單身聯誼,婚友,未婚聯誼,婚友社,未婚聯誼,單身聯誼,單身聯誼,婚友,單身聯誼,未婚聯誼,婚友,交友,交友,婚友社,婚友社,婚友社,大陸新娘,大陸新娘,大陸新娘,越南新娘,越南新娘,外籍新娘,外籍新娘,台中坐月子中心,搬家公司,搬家,搬家,搬家公司,線上客服,網頁設計,線上客服,網頁設計,網頁設計,土地貸款,免費資源,電腦教學,wordpress,人工植牙,關鍵字,關鍵字,seo,seo,網路排名,自然排序,網路排名軟體,
外遇外遇外遇外遇外遇外遇外遇外遇外遇 外遇
外遇 外遇外遇 外遇 外遇
外遇 外遇 外遇
外遇
外遇 外遇
外遇
外遇 外遇外遇
外遇
外遇 外遇外遇 外遇 外遇 , 外遇 , 外遇 , 外遇 , 外遇 , 外遇 , 外遇 , 外遇 , 外遇 , 外遇 ,
外遇 外遇 外遇 , 外遇 , 外遇 , 外遇 , 外遇 , 外遇 , 外遇 , 外遇 , 外遇
自慰,女生自慰,美女,高潮,失戀,性高潮,女性高潮,G點,自慰,性,性感,性感內衣,內衣,性感內衣,性感睡衣,情趣內衣,比基尼內衣,人妻少婦性感睡衣,
失戀,性愛,性感,情慾,調情,挑逗,慾望,情人節禮物,自慰器,
酒店經紀人,
菲梵酒店經紀,
酒店經紀,
禮服酒店上班,
酒店小姐兼職,
便服酒店經紀,
酒店打工經紀,
制服酒店工作,
專業酒店經紀,
合法酒店經紀,
酒店暑假打工,
酒店寒假打工,
酒店經紀人,
菲梵酒店經紀,
酒店經紀,
禮服酒店上班,
酒店經紀人,
菲梵酒店經紀,
酒店經紀,
禮服酒店上班,
酒店小姐兼職,
便服酒店工作,
酒店打工經紀,
制服酒店經紀,
專業酒店經紀,
合法酒店經紀,
酒店暑假打工,
酒店寒假打工,
酒店經紀人,
菲梵酒店經紀,
酒店經紀,
禮服酒店上班,
酒店小姐兼職,
便服酒店工作,
酒店打工經紀,
制服酒店經紀,
酒店經紀,
菲
梵,
Post a Comment