This case requires us to determine whether a gambling machine known as the Lucky Tab II, an electromechanical device that dispenses paper pull-tabs and then displays their contents on a video monitor, should be classified under the Indian Gaming Regulatory Act as a Class II "aid" or a Class III "facsimile." The Act prohibits Indian tribes from operating Class III facsimiles without first negotiating a compact with the state. Applying the statute's plain language, guided by our only relevant precedent, Cabazon Band of Mission Indians v. NIGC, 14 F.3d 633 (D.C. Cir. 1994), and proceeding without any views from the agency charged with the Act's implementation, we conclude that the Lucky Tab II is a Class II aid.

* The Indian Gaming Regulatory Act of 1988 ("IGRA"), 25 U.S.C. § 2701-19, regulates gambling operations run by Indian tribes. The Act's purpose is to "provide a statutory basis for the operation of gaming by Indian tribes as a means of promoting tribal economic development, self-sufficiency, and strong tribal governments." 25 U.S.C. § 2702(1).

The Act divides Indian gaming into three classes, each requiring a different level of authorization. Class I gaming consists of social games played solely for prizes of minimal value as well as traditional forms of Indian gaming. See 25 U.S.C. § 2703(6). Indian tribes may operate Class I games as they wish. See 25 U.S.C. § 2710(a)(1).

Class II gaming includes bingo, and if conducted in the same hall as bingo, it also includes lotto, punch boards, and tip jars, as well as pull-tabs, the game at issue here. See 25 U.S.C. § 2703(7)(A). In language central to the dispute in this case, the Act allows the use of "electronic, computer, or other technologic aids" in connection with Class II games, 25 U.S.C. § 2703 (7)(A)(i), but prohibits the use of "[e]lectronic or electromechanical facsimiles of any game of chance." 25 U.S.C. § 2703 (7)(B)(ii). Tribes may conduct Class II gaming if the state in which they are located permits such forms of gambling and if the governing body of the tribe adopts a gaming ordinance that is then approved by the Chairman of the National Indian Gaming Commission, the agency created by Congress to implement IGRA. See 25 U.S.C. §§ 2710(b), 2704.

Class III gaming includes all gambling not covered by either Class I or Class II, including "facsimiles" of Class II devices. See 25 U.S.C. § 2703(8). In order to conduct Class III operations, tribes must obtain state approval through negotiation of a tribal-state compact. See 25 U.S.C. § 2710(d)(1).

Commission regulations define Class II aids and Class III facsimiles. An aid is "a device ... that when used ... [i]s not a game of chance but merely assists a player or the playing of a game [and] is readily distinguishable from the playing of a game of chance on an electronic or electromechanical facsimile." 25 C.F.R. § 502.7. A facsimile is "any gambling device as defined in [the Johnson Act]." 25 C.F.R. § 502.8. Predating IGRA by more than 30 years, the Johnson Act prohibits the use of gambling devices on federal land, in interstate commerce, and in "Indian country." See 15 U.S.C. §§ 1171-78 (1953). Both the Commission's regulations and this Court have interpreted IGRA as limiting the Johnson Act prohibition to devices that are neither Class II games approved by the Commission nor Class III games covered by tribal-state compacts. See Cabazon, 14 F.3d at 635, n.3 (noting that IGRA repealed the Johnson Act with regard to Class III devices subject to a tribal-state compact but that there is no other repeal of the Johnson Act in IGRA, implying that Class II aids, permitted under IGRA, do not run afoul of the Johnson Act).

This case concerns a game known as pull-tabs. A small, two-ply paper card, a pull-tab bears symbols and patterns similar to tic-tac-toe that appear when players peel off the pull-tab's top layer. The pattern of the symbols determines whether the player wins a prize. In the traditional pull-tabs game, bingo hall clerks sell pull-tabs from counters or mobile carts, and winners present the tabs to either clerks or cashiers to collect prizes. Pull-tabs are sold from large pools known as "deals." Containing anywhere from 1200 to 100,000 pull-tabs, deals have a fixed number of winners and losers.

At issue in this case is the proper classification of a gambling device known as the Lucky Tab II, an electromechanical dispenser of paper pull-tabs. The machine dispenses pull-tabs from a roll containing approximately 7500 tabs.About 100 rolls comprise a deal, within which winning pulltabs are randomly distributed. The machine cuts the pull-tab from the roll and drops it into a tray. A bar code scanner inside the machine automatically reads the tab and then displays its contents on a video screen. A placard on the machine informs players that "[v]ideo images may vary from actual images on pull tabs. Each tab must be opened to verify." To collect prizes, players must present the actual winning tab to a clerk. In many bingo halls, players purchase pull-tabs either from a Lucky Tab II or from clerks; in such cases, machines and clerks cut pull-tabs from rolls that are part of the same deal.

In 1994, the Kickapoo Traditional Tribe of Texas and Diamond Game Enterprises, the manufacturer of the Lucky Tab II, asked the Commission to classify the machine as a Class II aid. Two years passed without Commission action.In August 1996, the Kickapoo Tribe began operating approximately 100 Lucky Tab II machines. At this point, the record becomes complicated and, to say the least, confusing. As far as we can tell, the following events of significance to this case transpired: The Commission's Director of Enforcement advised the Tribe that the machines were Class III gambling devices that could only be operated pursuant to a tribal-state compact. See Diamond Game Enterprises, Inc. v. Reno, 9 F. Supp. 2d 13, 15 (D.D.C. 1998). Notwithstanding the Director's action, the members of the Commission were apparently divided over the proper classification of the Lucky Tab II, some thinking it an aid and others a facsimile. Because of this disagreement, the Commission sought advice from the Department of Justice, but DOJ lawyers were themselves divided over the proper classification of the machine. See Memorandum from Deputy Assistant Attorney General Richard Shiffrin to Associate Deputy Attorney General Seth P. Waxman, at 1 (June 13, 1996) (noting that the Office of Tribal Justice and the Criminal Division had reached opposite conclusions on the appropriate classification of the Lucky Tab II--the former concluding that it falls under Class II and the latter concluding that it belongs in Class III). The Commission never formally responded to the request to classify the Lucky Tab II.

According to the Tribe and Diamond Game, certain members of the Commission recommended that the Tribe and the company file a declaratory judgment action in federal court to resolve the issue. Acting on that advice, they filed this action in the U.S. District Court for the District of Columbia seeking, among other things, a declaratory judgment that the machine qualifies as a Class II aid. The Cheyenne and Arapaho Tribes of Oklahoma intervened as plaintiffs. Alabama, California, and Florida intervened as defendants.

The parties filed cross motions for summary judgment.Finding that the Lucky Tab II "performs all the functions that a player of the traditional pull-tab game would have performed," the district court found the machine to be a Class III facsimile and granted summary judgment to the government. See Diamond Game, 9 F. Supp. 2d at 20. Subsequently, Diamond Game and the Tribes filed a Rule 60(b) motion, claiming that the company had made technical changes to the Lucky Tab II. Finding that the modifications were not new evidence, the district court denied the motion.

Unlike the legal issues presented in this case, the policy questions are both interesting and challenging. In determining the proper classification of the Lucky Tab II, how do we further Congress' objective of allowing Indian tribes to use gaming as a means of "promoting tribal economic development, self-sufficiency, and strong tribal governments," 25 U.S.C. § 2702(1), while at the same time "shield[ing] [Indian tribes] from organized crime and other corrupting influences," 25 U.S.C. § 2702(2)? Will the Lucky Tab II enable tribes to "take advantage of modern methods of conducting class II games"? S. Rep. No. 100-446, at 9 (1988). Or does the machine increase the risk of corruption or excessive gambling losses, concerns that government counsel told us at oral argument require its classification as a Class III device? To resolve such issues, Congress created the National Indian Gaming Commission, headed by a Chair appointed by the President and confirmed by the Senate presumably for his or her expertise on Indian gaming. Yet whether because of bureaucratic gridlock or, as the tribes allege, because of congressional interference, we have no idea what the Commission thinks about the policy questions presented by the Lucky Tab II. Not only does this leave us with no agency position to which we might defer, see Chevron U.S.A. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 844 (1984) ("[A] court may not substitute its own construction of a statutory provision for a reasonable interpretation made by the administrator of an agency"), but the Commission's IGRA regulations provide no assistance in interpreting the statute. Boiled down to their essence, the regulations tell us little more than that a Class II aid is something that is not a Class III facsimile. We mention this not to escape our duty to decide this case--to the contrary, because we have jurisdiction, we must determine how the Lucky Tab II should be classified--but to highlight the fact that we have no choice but to proceed without the benefit of a Commission position, a situation we expect Congress neither anticipated nor would appreciate. That said, we turn to the parties' arguments about the classification of the Lucky Tab II. See Everett v. United States, 158 F.3d 1364, 1367 (D.C. Cir. 1998) ("We review a grant of summary judgment de novo.").

Diamond Game and the Tribes contend that the Lucky Tab II acts as a permitted "electronic aid" to the Class II game of pull-tabs. They emphasize that the machine's operation depends entirely on pre-printed paper pull-tabs that can be (and in fact are) played without the mechanical dispenser. The Lucky Tab II, in other words, cannot function without rolls of paper pull-tabs. The Tribes also emphasize that despite the fact that the Lucky Tab II presents a video image of the contents of the pull-tabs it dispenses, the machine does not give the player the final word on the game; players must still peel off the top layer to verify its contents and present it to a clerk to receive their winnings. For all of these reasons, they argue, the Lucky Tab II cannot be considered a facsimile of the paper game of pull-tabs.

According to the government, because the machine mirrors the traditional game played by purchasing cards from clerks, it is a Class III facsimile, not a Class II aid. The government embraces the district court's description of the Lucky Tab II:"When the participant plays the Lucky Tab II, she is not playing the pull-tabs inside the machine; she is engaging the machine that replicates the functions of the traditional pulltab game." Diamond Game, 9 F. Supp. 2d at 13, 20. As to the possibility that the information on the video screen might be inaccurate, the government says mistakes are rare and for all practical purposes, the Lucky Tab II is a duplicate of the paper version.

Both sides claim support from Cabazon Band of Mission Indians v. NIGC, 14 F.3d 633 (D.C. Cir. 1994). There, we held that a video pull-tabs game was a "computerized version" of pull-tabs and therefore a Class III facsimile. The machine "randomly selects a card for the gambler, pulls the tab at the gambler's direction, and displays the result on the screen. The computer version, like the paper version, has a fixed number of winning cards in each deal." Cabazon, 14 F.3d at 635. Finding that video pull-tabs "exactly replicate[s]" the game of pull-tabs in computer form, Cabazon concluded that it amounted to a facsimile of the game. See id.; see also Sycuan Band of Mission Indians v. Roache, 54 F.3d 535, 541-42 (9th Cir. 1994) (holding that a self-contained unit containing a computer linked to a video monitor and a printer constitutes an electronic facsimile of pull-tabs).

We think the Lucky Tab II is quite different from the machine at issue in Cabazon. To begin with, the Lucky Tab II is not a "computerized version" of pull-tabs. Although the Lucky Tab II has a video screen, the screen merely displays the contents of a paper pull-tab. Instead of using a computer to select patterns, the Lucky Tab II actually cuts tabs from paper rolls and dispenses them to players. In other words, the game is in the paper rolls, not, as in the case of the Cabazon machine, in a computer. Indeed, players using the Lucky Tab II often play a deal simultaneously with other players in the same hall who have chosen to purchase pull tabs from clerks. For players using the Lucky Tab II, the machine functions as an aid--it "helps or supports," or "assists" the paper game of pull-tabs. Webster's Third New International Dictionary 44 (1993). Without the paper rolls, the machine has no gaming function at all. It is, in essence, little more than a high-tech dealer. Viewed this way, the game played with the Lucky Tab II is not a facsimile of paper pull-tabs, it is paper pull-tabs.

Another difference between the Lucky Tab II and the video machine at issue in Cabazon reinforces our belief that the Lucky Tab II should be classified as a Class II aid. The Cabazon machine plays the game of pull-tabs in its entirety, dispensing receipts for players to redeem winnings. By contrast, the Lucky Tab II dispenses actual paper pull-tabs that players must peel and display to a clerk before they can obtain prizes. Although the machine's scanner apparently commits few errors when reading paper pull-tabs, the fact remains that unlike the Cabazon machine, the Lucky Tab II is technically not final. It is, in other words, an aid to the game of pull-tabs.

Notwithstanding the differences between the Lucky Tab II and the machine at issue in Cabazon, the government insists that the Lucky Tab II is a Class III device. At oral argument, the government even asserted that removing the video screen would not convert the Lucky Tab II into a Class II aid. Asked what in the government's view would be an aid, counsel pointed us to an electronic scanner called the "Tab Force Validation System." As we understand this device, after a clerk dispenses a paper pull-tab, instead of peeling off the top layer, the player inserts the pull-tab into the machine, which scans the bar code and displays the results on a video screen. The Commission has issued advisory opinions classifying the Tab Force and other similar machines as Class II aids, concluding that the systems "simply read the pull-tabs and display whether or not they are winners.... [They] cannot change the outcome of the game." See NIGC Advisory Opinion, at 2 (June 8, 1998).

We see no principled difference between the Tab Force and the Lucky Tab II. Both devices electronically "read" paper pull-tabs and display their contents on a screen, and neither can "change the outcome of the game." Unlike the machine involved in Cabazon, neither contains an internal computer that generates the game. Rather, both machines facilitate the playing of paper pull-tabs. They are thus Class II aids.

The government makes two additional arguments in support of its position that the Lucky Tab II is a Class III facsimile. First, like the district court, it relies on language from a Senate Indian Affairs Committee report describing a Class II aid as a device that enables tribes to "take advantage of modern methods of conducting class II games" by, for example, "join[ing] with other tribes to coordinate their class II operations and thereby enhance the potential of increasing revenues." S. Rep. No. 100-446, at 9 (1988). Class II aids are thus limited to devices that "merely broaden the potential participation levels and [are] readily distinguishable from ... electronic facsimiles in which a single participant plays a game with or against a machine rather than with or against other players." Id. Unlike computers, cables, or telephone lines that connect bingo games on different reservations-examples the Senate Report gives of aids that expand participation--the Lucky Tab II, the government argues, neither increases participation levels nor enhances competition among players. Second, the government claims that the Lucky Tab II makes it easier for players to play pull-tabs, thus increasing the potential for players to "lose the rent money."

These statutory interpretations, resting as they do on the policy underlying IGRA, are interesting and might even be worthy of Chevron two deference had they been offered by the Commission. But they come only from appellate counsel--indeed the "lose-the-rent" argument surfaced for the first time at oral argument. Moreover, nothing in the Senate Report suggests that an electronic device must link players on different reservations to qualify as a Class II aid. Accordingly, because of the similarities between the Lucky Tab II and the Tab Force Validation System, which the Commission has found to be a Class II aid, and because of the differences between the Lucky Tab II and the Class III device at issue in Cabazon, we reverse the district court and remand with instructions to enter summary judgment for appellants. In view of this disposition, we have no need to address the district court's denial of the Rule 60(b) motion.